FINITE ELEMENT MODELING OF LARGE DEFORMATION
RESPONSE OF REINFORCED CONCRETE BEAMS
A Thesis Presented
by
Andre Werner
To
The Department of Civil and Environmental Engineering
in partial fulfillment of the requirements
for the degree of
Masters of Science
in
Civil Engineering
in the field of
Structural Engineering
Northeastern University
Boston, MA
August 2008
NORTHEASTERN UNIVERSITY
GRADUATE SCHOOL OF ENGINEERING
Thesis Title: Finite Element Modeling of Large Deformation Response of
Reinforced Concrete Beams
Author: Andre Werner
Department: Civil and Environmental Engineering
Approved for Thesis Requirement of the Master of Science Degree
_____________________________________________________ ____________ Thesis Advisor: Mehrdad Sasani, Ph.D., P.E. Date
_____________________________________________________ ____________ Thesis Reader: Dionisio Bernal, P.h.D Date
_____________________________________________________ ____________ Department Chair: Thomas C. Sheahan, Sc.D, P.E. Date
Graduate School Notified of Acceptance
_____________________________________________________ ____________ Director of the Graduate School: Yaman Yener, Ph.D Date
i
ABSTRACT
In a previous study, a 3/8 scaled physical model of a continuous two-span reinforced
concrete (RC) beam was constructed and tested. The RC beam was subjected to a
downward vertical displacement. In this study a detailed three dimensional finite element
model of the RC beam is developed. Unlike a previous analytical study, the contribution
of bond to the response of the beam is considered here by deploying interface elements
between longitudinal reinforcement elements and concrete elements of the finite element
model. Nonlinear material behavior such as concrete cracking and crushing and steel
reinforcement yielding are also modeled. A geometrically nonlinear incremental static
analysis using the computer program DIANA 9.2 is carried out. The model is analyzed
under the applied deformations at the centerline. Analysis results include element forces,
stresses and strains of the concrete and steel reinforcement and bond stresses and slippage
along the beam. Stress-strain relationships of steel reinforcement and beam end rotations
are compared with the corresponding experimental results. Overall, the analytical and
experimental results are in good agreement.
Table of Contents
Abstract i
Table of Contents ii
List of Figures iv
List of Tables viii
Acknowledgements ix
1 Introduction 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Structure of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Spatial Discretization and Element Topology 3
2.1 Cross Section Discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 3-D Discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1 Interface between Longitudinal Reinforcement and Concrete . . . . . . 6
2.2.2 Transverse Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . 6
3 Material Constitutive Models 7
3.1 Steel Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.1 Theoretical Description . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.2 Material Parameters of Steel Reinforcement . . . . . . . . . . . . . . . 9
3.2 Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.1 2-D Modified Compression Field Theory . . . . . . . . . . . . . . . . 10
CONTENTS iii
3.2.2 3-D Modified Compression Field Theory . . . . . . . . . . . . . . . . 12
3.2.3 Concrete in Tension . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2.4 Concrete Material Parameters . . . . . . . . . . . . . . . . . . . . . . 16
Concrete Compressive Stress-Strain Relationships . . . . . . . . . . . 16
Concrete Tensile Stress-Strain Relationship . . . . . . . . . . . . . . . 19
3.3 Interface between Concrete and Steel Reinforcement . . . . . . . . . . . . . . 21
3.3.1 Mechanics of Interaction between Steel Reinforcement and Concrete . 21
3.3.2 Review of Existing Bond Stress-Slip Relationships . . . . . . . . . . . 23
3.3.3 Parameters influencing Bond Performance . . . . . . . . . . . . . . . . 27
3.3.4 Implemented Bond Stress-Slip Relationships . . . . . . . . . . . . . . 29
4 Analytical Results 35
4.1 Analytical Results at Material Level . . . . . . . . . . . . . . . . . . . . . . . 36
4.2 Bond Slip and Section Deformation . . . . . . . . . . . . . . . . . . . . . . . 47
4.3 Element End Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4 State of Strain at Beam Ends . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5 Comparison between Analytical and Experimental Results 60
5.1 Steel Reinforcement Strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.2 Beam End Rotations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6 Concluding Remarks 63
References 65
List of Figures2.1 View of the RC beam reported in [1] . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Cross section A-A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 Cross section B-B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.4 Discretization of the cross section of the RC beam reported in [1] . . . . . . . . 4
2.5 3-D finite element model of the RC beam reported in [1] . . . . . . . . . . . . 5
2.6 Topology of solid element . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.7 Topology of interface element and associated nodal displacements . . . . . . . 6
3.1 Uniaxial stress-strain relationships of longitudinal and transverse steel rein-
forcements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 2-D state of stress in a reinforced concrete element [7] . . . . . . . . . . . . . 11
3.3 2-d state of strain in a reinforced element [7] . . . . . . . . . . . . . . . . . . . 11
3.4 Uniaxial compressive stress-strain relationship for concrete cover, side support
and center column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.5 Effectively confined concrete core . . . . . . . . . . . . . . . . . . . . . . . . 19
3.6 Uniaxial compressive stress-strain relationship for concrete core [11] and the
trilinear approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.7 Uniaxial tensile stress-strain relationship of reinforced concrete [10] . . . . . . 20
3.8 Qualitative local bond stress-slip relationships [13] . . . . . . . . . . . . . . . 21
3.9 Transverse cracks originated at the tips of the steel reinforcement ribs and
crushed concrete in front of the ribs of the rebar [12] . . . . . . . . . . . . . . 22
3.10 Radial component of bond stress balanced by tensile ring stresses in the sur-
rounding concrete [14] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.11 Radial cracks induced by tensile ring stresses and transverse reinforcement ac-
tion [13] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.12 Bond stress-slip relationships [12], [15] and [17] . . . . . . . . . . . . . . . . 27
LIST OF FIGURES v
3.13 Defined regions of bond in the finite element model . . . . . . . . . . . . . . . 30
3.14 Pattern of the bond stress-slip relationship in region A and B . . . . . . . . . . 30
3.15 Hydraulic pressure analogy taken from [14] . . . . . . . . . . . . . . . . . . . 31
3.16 Bond stress-slip relationship of interface elements in region A and B as imple-
mented in the finite element model . . . . . . . . . . . . . . . . . . . . . . . . 34
4.1 Designation of longitudinal views used for contour plots . . . . . . . . . . . . 35
4.2 Locations of concrete C, steel S and interface I elements . . . . . . . . . . . . 35
4.3 Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at
0.15, 0.3 and 0.48 in of vertical displacement . . . . . . . . . . . . . . . . . . 37
4.4 Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at
1.45, 1.68 and 1.9 in of vertical displacement . . . . . . . . . . . . . . . . . . 37
4.5 Longitudinal profile 1 of first principal stress σc,1 [psi] at three vertical dis-
placements: 0.76 (Top); 1.0 (Middle) and 1.2 in (Bottom) . . . . . . . . . . . . 38
4.6 Variation of concrete longitudinal stress σc,zz at cross sectional location C1 at
1.45, 1.68 and 1.9 in of vertical displacement . . . . . . . . . . . . . . . . . . 39
4.7 Variation of concrete longitudinal stress σc,zz at cross sectional location C2 at
1.45, 1.68 and 1.9 in of vertical displacement . . . . . . . . . . . . . . . . . . 39
4.8 Variation of concrete longitudinal stress σc,zz at cross sectional location C4 at
0.76, 1.0 and 1.2 in of vertical displacement . . . . . . . . . . . . . . . . . . . 40
4.9 Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at
0.76, 1.0 and 1.2 in of vertical displacement . . . . . . . . . . . . . . . . . . . 40
4.10 Longitudinal profile 1 of third principal stress σc,3 at 1.2, 1.45, 1.68 and 1.9 in
(plots from top to bottom) of vertical displacement (the face of the side support
is located on the left, the face of the center column is located on the right) . . . 41
4.11 Longitudinal profile 2 of third principal stress σc,3 at 1.2, 1.45, 1.68 and 1.9 in
(plots from top to bottom) of vertical displacement (the face of the side support
is located on the right, the face of the center column is located on the left) . . . 41
4.12 Variation of steel longitudinal strain at cross sectional location S1 at 4.44 and
5.26 in of vertical displacement (horizontal dashed lines represent yield strains) 42
4.13 Variation of steel longitudinal stress at cross sectional location S1 at 4.44 in and
5.26 in of vertical displacement (horizontal dashed lines represent yield stresses) 43
4.14 Variation of steel longitudinal strain at cross sectional location S2 at 4.44 and
5.26 in of vertical displacement (horizontal dashed lines represent yield strains) 43
LIST OF FIGURES vi
4.15 Tensile stress-strain of the first three steel reinforcement elements from the
face of the center column at cross sectional location S2. (Uniaxial stress-strain
relationship of steel reinforcement as implemented in the finite element model
is also shown.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.16 Variation of bond slip at cross sectional location I1 at 1.45, 1.68 and 1.9 in of
vertical displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.17 Variation of bond stress at cross sectional location I1 at 1.45, 1.68 and 1.9 in of
vertical displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.18 Variation of bond slip at cross sectional location I2 at 2.1, 2.35 and 2.58 in of
vertical displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.19 Variation of bond stress at cross sectional location I2 at 2.13, 2.35 and 2.58 in
of vertical displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.20 Designation of beam sections . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.21 Deformation [in] of SEC 2 at vertical displacements of 1.0 in (top) and 1.9 in
(bottom) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.22 Deformation [in] of SEC 3 at vertical displacements of 1.0 in (top) and 1.9 in
(bottom) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.23 Deformation [in] of SEC 1 at vertical displacement of 1.0 in . . . . . . . . . . 50
4.24 Deformation [in] of SEC 4 at vertical displacement of 1.0 in . . . . . . . . . . 50
4.25 Deformation [in] of SEC 1 at vertical displacement of 1.9 in . . . . . . . . . . 50
4.26 Deformation [in] of SEC 4 at vertical displacement of 1.9 in . . . . . . . . . . 51
4.27 Moment at the boundary of the center column vs. imposed vertical displacement 52
4.28 Moment at the boundary of the side support vs. imposed vertical displacement . 53
4.29 Shear force response vs. imposed vertical displacement . . . . . . . . . . . . . 53
4.30 Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 0.15,
0.3 and 0.48 in of vertical displacement (horizontal dashed line represents yield
strain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.31 Variation of concrete longitudinal strain εc,zz at cross sectional location C1 at
0.15, 0.3 and 0.48 in of vertical displacement (horizontal dashed lines represent
peak compressive strain and crack strain) . . . . . . . . . . . . . . . . . . . . 55
4.32 Variation of concrete longitudinal strain εc,zz at cross sectional location C4 at
0.76, 1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent
peak compressive strain and crack strain) . . . . . . . . . . . . . . . . . . . . 56
LIST OF FIGURES vii
4.33 Variation of steel longitudinal strain εs,zz at cross sectional location S1 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed line represents yield
strain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.34 Variation of concrete longitudinal strain εc,zz at cross sectional location C2 at
0.76, 1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent
peak compressive strain and crack strain) . . . . . . . . . . . . . . . . . . . . 57
4.35 Variation of concrete longitudinal strain εc,zz at cross sectional location C1 at
0.76, 1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent
peak compressive strain and crack strain) . . . . . . . . . . . . . . . . . . . . 57
4.36 Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed line represents yield
strain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.37 Variation of concrete longitudinal strain εc,zz at cross sectional location C3 at
0.76, 1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent
peak compressive strain and crack strain) . . . . . . . . . . . . . . . . . . . . 58
4.38 Variation of steel longitudinal strain εs,zz at cross sectional location S1 at about
1.45, 1.68 and 1.9 in of vertical displacement (horizontal dashed lines represent
yield strains) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.39 Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 1.45,
1.68 and 1.9 in of vertical displacement (horizontal dashed lines represent yield
strains) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.1 Analytical and experimental strain history of bottom rebar at the face of the
center column (SEC 4) vs. imposed vertical displacement . . . . . . . . . . . . 60
5.2 Analytical and experimental strain history of top rebar at the face of the side
support (SEC 1) vs. imposed vertical displacement . . . . . . . . . . . . . . . 62
5.3 Analytical and experimental beam rotations over 8 in from the face of center
column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.4 Analytical and experimental beam rotations over 8 in from the face of the side
support over a length of 8 in . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
List of Tables3.1 Material parameters and characteristic values of uniaxial stress-strain relation-
ships of longitudinal and transverse steel reinforcement . . . . . . . . . . . . . 9
3.2 Material parameters and characteristic values of uniaxial compressive stress-
strain relationship of concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.3 Parameters of tensile stress-strain relationship for reinforced concrete . . . . . 20
3.4 Initial characteristic values of the bond stress-slip relationships in region A and B 30
3.5 Characteristic values of interface bond stress-slip relationships of region A and
B as implemented in the finite element model . . . . . . . . . . . . . . . . . . 33
4.1 Analysis load steps and associated vertical displacement at the face of center
column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2 Values of steel reinforcement pull-out and push-in at about 5.26 in of vertical
displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
ACKNOWLEDGMENTS
I would like to express my gratitude to my advisor, Professor Mehrdad Sasani, for his
guidance, patience and invaluable support throughout this study. I would like to thank
Professor Dionisio Bernal for kindly accepting to read my thesis. I would also like to
thank Professor Peter Furth who supported my aspiration to study at Northeastern
University. I also wish to express my appreciation to my parents for their support and
encouragement.
.
Chapter 1 Introduction
1.1 Overview
Understanding the behavior of reinforced concrete (RC) beams subjected to large deformations
requires further analytical and experimental studies. In [1], analytical and experimental studies
of an RC beam subjected to large deformations are presented using a detailed 3-D finite element
model. Yet, the bond slip between steel reinforcement was not modeled. In this study, a
different finite element model of an RC beam is developed, incorporating bond slip.
1.2 Objectives
The goals of this thesis are:
1) To develop a detailed three dimensional finite element model of a RC beam to be subjected
to large deformations.
2) To evaluate the model by comparing the analytical results with the previously obtained
experimental data.
1.3 Structure of the Thesis
In chapter 2, the spatial discretization of the RC beam is discussed. The topology of the el-
ements used to model the concrete core and concrete cover, longitudinal and transverse steel
reinforcement as well as the interface between concrete and steel are introduced.
In chapter 3, the constitutive modeling of concrete, steel reinforcement and interface steel-
concrete is discussed. Using von Mises yield criteria, input stress-strain relationships for the
transverse and longitudinal steel reinforcement are introduced. Utilizing the Modified Com-
pression Field Theory, material modeling for concrete in compression and tension is presented.
Evaluating existing bond-slip relationships and the corresponding experimental data, input
bond-slip relationships for different regions in the RC beam are presented.
INTRODUCTION 2
In chapter 4, the results of the analysis are presented. The variation of stresses and strains
in reinforcement steel and concrete as well as of bond stresses and slippage along the beam
is shown. The deformations of the RC beam sections are discussed and pull-out and push-in
values of the anchored steel reinforcement are presented. The response of the RC beam at ma-
terial, section and element level is presented.
In chapter 5, analytical steel reinforcement strains are evaluated and compared with the corre-
sponding experimentally results. The analytical beam end rotations are also compared with the
experimental results.
In chapter 6, concluding remarks are presented.
Chapter 2 Spatial Discretization and Element
Topology
Figs. 2.1 to 2.3 show an RC beam reported in [1] and its cross sections. In order to capture
the three dimensional state of stress and strain in the RC beam, a 3-D finite element model is
developed. The interface between concrete and steel reinforcement is explicitly modeled using
interface elements.
Figure 2.1: View of the RC beam reported in [1]
Figure 2.2: Cross section A-A Figure 2.3: Cross section B-B
SPATIAL DISCRETIZATION AND ELEMENT TOPOLOGY 4
2.1 Cross Section Discretization
The discretization of beam section A-A is shown in Fig. 2.4. Concrete cover and core main
areas are built of quadrilateral surfaces. The steel reinforcement cross sections comprise oc-
tagonal surfaces. The transition region from octagonal to quadrilateral surfaces is discretized
using triangular surfaces. Given the symmetric cross section of the beam, only half of the beam
cross section is modeled.
Figure 2.4: Discretization of the cross section of the RC beam reported in [1]
2.2 3-D Discretization
In order to save computational time, only a quarter of the RC beam is modeled. That is, one half
of the cross section and one half of the two span beam is modeled. The finite element model
of the experiment reported in [1] is composed of 5 Segments (see Figs. 2.1, 2.2, 2.3 and 2.5).
Segment 1 includes a portion of the beam having the cross section shown in Fig. 2.3. Segments
2 and 3 have the cross section shown in Fig. 2.2. Segment 4 is the side support of the RC beam
and its right hand side boundary nodes are fixed in all three translational degrees of freedom.
Segment 5 is the center column part of the RC beam. At the left hand side boundary of segment
5, the experimental measured RC beam deformations will be imposed to the finite element
model. Transverse reinforcement is modeled using the embedded reinforcement concept which
is described later in this chapter. The top longitudinal reinforcement of the tested RC beam was
anchored in the side support using a 90 degree hook. The top reinforcement of the finite element
3-D DISCRETIZATION 5
model of the RC beam is anchored straight over a length of 11.0 in within the side support part
in order to minimize discretization effort. Only small values of steel reinforcement longitudinal
stress and strain are present at that location. Thus, it is believed that omitting to model the
anchorage hook in the model will not significantly influence the results of the computation.
20-node isoparametric brick elements are selected to model the concrete cover and core (see
Fig. 2.6). The longitudinal steel reinforcement of the RC beam is also modeled using 20-node
isoparametric brick elements. This was primarily because of the need for an element to be
compatible with the available interface elements in program DIANA 9.2 as discussed in the
next section.
X
Y
Z
Figure 2.5: 3-D finite element model of the RC beam reported in [1]
1
2
3 4
5
678
9
10
11
1213 14
15
16 17
1819
20
ξ
η
ζ
Figure 2.6: Topology of solid element
SPATIAL DISCRETIZATION AND ELEMENT TOPOLOGY 6
2.2.1 Interface between Longitudinal Reinforcement and Concrete
The area between longitudinal steel reinforcement and concrete is modeled using 16-node plane
quadrilateral isoparametric elements providing the capability to discretize curved surfaces (Fig.
2.7). The element comprises two overlaying planes. Each pair of nodes have coincident coor-
dinates which implies zero thickness of the interface element.
1 2
3
45
67
8
x
yz9 10
11
1213
1415
16
ux
uyuz
Figure 2.7: Topology of interface element and associated nodal displacements
The set of variables describing the element state of deformation and the element forces contains
nodal displacements ue, relative displacements Δu as well as normal traction tx and shear
tractions ty and tz. The location of the element nodes and the variables of the chosen interface
element are shown in Fig.2.7.
ue =
⎛⎜⎝
ux
uy
uz
⎞⎟⎠ t =
⎛⎜⎝
tx
ty
tz
⎞⎟⎠ Δu =
⎛⎜⎝
Δux
Δuy
Δuz
⎞⎟⎠ (2.1)
2.2.2 Transverse Reinforcement
Transverse reinforcement is modeled using embedded reinforcement. The strain of the rein-
forcement bar is computed from the displacement field of the concrete element in which the
transverse reinforcement is embedded.
Chapter 3 Material Constitutive Models
3.1 Steel Reinforcement
3.1.1 Theoretical Description
The use of solid elements to discretize the longitudinal steel reinforcement of the RC beam
(Section 2.2) requires a yield criteria to define the elastic limit of the steel reinforcement under
combined state of stresses. The DIANA 9.2 program provides the Tresca and von Mises yield
criteria which are both isotropic and independent of hydrostatic pressure and therefore suitable
for modeling steel reinforcement represented by means of solid elements. Experimental tests
on ductile steel have shown a particulary good agreement with von Mises criteria in describing
the onset of yielding not only under one-dimensional state of stress but also under combined
state of stresses [2]. Thus, the von Mises yield criteria is used in this study.
The von Mises yield criteria states that the onset of yielding depends on whether the octahedral
shear stress or the maximum distortional energy has reached a critical value [3]
f(σij, κ) = J2 − κ2 = 0 (3.1)
where κ represents the yield stress in pure shear; σij represents the state of stress and J2 de-
notes the second deviatoric stress invariant.
In the literature, Eq. 3.1 is also referred to as yield surface in principal stress space [3]. If the
von Mises criteria is calibrated to agree with the uniaxial tension yield stress σyield, κ is
κ =σyield√
3(3.2)
In program DIANA 9.2, the von Mises yield criteria is implemented such that it agrees with the
uniaxial stress-strain relationship of steel reinforcement. Eq. 3.3 represents the von Mises yield
surface in principal stress space calibrated to uniaxial tension test as it is available in program
DIANA 9.2. It can be obtained by substituting Eq. 3.2 into Eq. 3.1.
MATERIAL CONSTITUTIVE MODELS 8
f(σij, σyield) =√
3J2 − σyield = 0 (3.3)
If the von Mises yield criteria in Eq. 3.3 is used to describe steel reinforcement material behav-
ior, only elastic-perfectly plastic material behavior can be modeled since plastic deformation
are assumed to occur under a constant yield stress. Yet, steel reinforcement exhibits hardening
behavior. In order to model this post-yield response, a hardening rule is used within the theory
of incremental plastic flow. A hardening rule specifies the configuration of subsequent yield
surfaces in stress space if a material is loaded beyond the initial yield surface represented by
Eq. 3.3. Hardening rules can be categorized in isotropic, kinematic and mixed hardening rules.
In Program DIANA 9.2 an isotropic hardening rule is implemented in conjunction with the
von Mises yield criteria. It is adopted in this study to model the hardening behavior of steel
reinforcement and it is described subsequently.
Eq. 3.4 represents the von Mises yield criteria presented in Eq. 3.3 extended by an isotropic
hardening rule [4].
f(σij, σyield, εp) =√
3J2 − σyield(εp) = 0 (3.4)
εp (effective strain [3]), is an increasing function of the accumulated plastic strain increments
dεpij . There are two basic approaches within the theory of plastic flow to define the effective
strain εp, which are work and strain hardening approaches. In the Diana 9.2 program both
approaches are implemented. Yet, in case of von Mises isotropic hardening material, both lead
to the same scalar function for εp [3]. Within the strain hardening approach, the effective strain
can be defined as a suitable combination of plastic strain increments dεpij [3]. In the DIANA
9.2 program the following definition is implemented.
εp =
∫ √2
3dεp
ijdεpij (3.5)
Having εp, σyield(εp) can be found using uniaxial material test results. The corresponding
associated flow rule implemented in the program DIANA 9.2 is used in this study.
STEEL REINFORCEMENT 9
3.1.2 Material Parameters of Steel Reinforcement
Experimental test results [1] as well as the data provided in [5] are used to determine the ma-
terial parameters for longitudinal and transverse reinforcements, which are presented in Table
3.1. The modulus of elasticity E is taken from [5]. The yield stress σyield for both reinforce-
ments has been determined from tension tests conducted in [1]. For the longitudinal steel
reinforcement of the RC beam a trilinear idealization of the experimental uniaxial stress-strain
relationship is adopted. For transverse reinforcement a bilinear idealization of the experimental
uniaxial stress-strain relationship is used. Both uniaxial stress-strain relationships are shown in
Fig. 3.1.
Table 3.1: Material parameters and characteristic values of uniaxial stress-strain relationships
of longitudinal and transverse steel reinforcement
Longitudinal Reinforcement Transverse Reinforcement
E [ksi] 29,000 29,000
σyield [ksi] 75 60
εyield 0.0026 0.0021
Epostyield [ksi] 287 306
μ 0.3 0.3
σult = [ksi] 105 90
εult 0.13 0.13
Figure 3.1: Uniaxial stress-strain relationships of longitudinal and transverse steel reinforce-
ments
MATERIAL CONSTITUTIVE MODELS 10
3.2 Concrete
In program DIANA 9.2, the three approaches to establish constitutive relations for concrete in
compression are: plasticity, elasticity and damage based approach.
Three-dimensional material models based on plasticity approach, namely the Drucker-Prager
model and the Modified Mohr-Coulomb model, are available in program DIANA 9.2. Within
the plasticity approach the total concrete strain εc is decomposed in an elastic part εpc and in a
plastic part εpc . The concept of strain decomposition appeals the most to the physical nature of
cracked concrete especially if combined with the smeared crack approach where the total strain
is decomposed to the ”crack strain” and the strain of the solid concrete [6]. Yet, preliminary
simulations using Drucker-Prager model in compression combined with a multi-directional
smeared crack model in tension revealed numerical difficulties (convergence problems). Ac-
cording to the manual of program DIANA 9.2 [4], such difficulties are likely to occur when in
one integration point simultaneously concrete compressive softening is accompanied by sev-
eral active cracks in the remaining lateral directions.
An elasticity based concept is the Modified Compression Field Theory. It was originally
developed within a two-dimensional framework [7]. The purpose was to predict the load-
deformation response of plane cracked reinforced concrete elements subjected to in-plane shear
and normal stresses. Subsequently, the concept was extended and enhanced, respectively, in or-
der to describe the load-deformation characteristic of reinforced concrete solids subjected to a
general three-dimensional state of stress [8]. This enhanced three-dimensional model con-
stitutes the basis for the elasticity based total strain concrete material model available in the
program DIANA 9.2.
Lastly, a damage based material model, namely the Modified Maekawa Model, is implemented
in the program DIANA 9.2. This model is based on two-dimensional and three-dimensional
cyclic loading data. The damage based approach was not considered in this study.
3.2.1 2-D Modified Compression Field Theory
Using a total strain based approach, concrete behavior in compression and tension is modeled
in this study. The total strain based approach for concrete material modeling in the program DI-
ANA 9.2 is based on the Modified Compression Field Theory [7]. In the Modified Compression
CONCRETE 11
Field Theory, stress-strain relationships are evaluated in terms of average stresses and strains.
Figs. 3.2 and 3.3 show a reinforced concrete element and the corresponding two-dimensional
states of stress and strain.
Figure 3.2: 2-D state of stress in a re-
inforced concrete element [7]
Figure 3.3: 2-d state of strain in a re-
inforced element [7]
In order to relate the principal average stresses σc,1 and σc,2 to principal average strains ε1 and
ε2 , an average stress-strain constitutive relationship is required. Such an average stress-strain
relationship can differ from local stress-strain relationships determined from standard material
tests.
An extensive series of reinforced concrete panels were tested in [7] in order to obtain infor-
mation with regard to characteristics of such an concrete average stress-strain constitutive re-
lationship. For details with regard to the set up the reader is referred to [7]. In these tests,
in-plane normal stresses σx , σy and shear stresses τxy, depicted in Fig. 3.2, were applied to the
reinforced concrete panels. The corresponding average strains εx , εy and γxy, shown in Fig.
3.3, were measured. Based on these measured quantities, the following average stress-strain
constitutive relationship for concrete in compression was developed:
σc2 = σc2,max ·[2
(ε2
ε′c
)−(
ε2
ε′c
)2]
(3.6)
where σc2 represents the second principal stress in concrete (maximum compression); ε2
represents the second principal strain in concrete and
MATERIAL CONSTITUTIVE MODELS 12
σc2,max
f ′c
=1
0.8 − 0.34 · ε1/ε′c(3.7)
where f ′c represents the concrete compressive strength; ε′c represents the concrete strain at f ′
c
and ε1 represents the first principal strain in concrete.
Eq. 3.6 and 3.7 reflects the main findings of the study with regard to the compressive load-
deformation characteristics of reinforced concrete panels. As one can see, looking at the con-
stitutive relationships presented above, the principal average compressive stress σc2 is not only
a function of the principal compressive strain ε2 but it also depends on the principal tensile
strain ε1. In other words, if cracked concrete is subjected to high tensile strains normal to the
compressive direction, it shows softer and weaker characteristic in compression.
3.2.2 3-D Modified Compression Field Theory
Section 3.2.1 was focussed on the load-deformation response of plane reinforced concrete el-
ements subjected to plane state of stress and strain incorporating compressive strength degra-
dation of concrete if tensile strain in the lateral direction prevails. However, the enhancement
of compressive strength and ductility of concrete due to confining effects has not been consid-
ered. The 3-D extension of the Modified Compression Field Theory [8] and the corresponding
proposed orthotropic secant material matrix for a concrete solid element, summarized in the
following section, accounts for the aforementioned concrete material behavior.
The three dimensional secant material matrix of an orthotropic concrete material in principal
directions is [8]
[D] =1
φ
⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣
Ec1(1 − υ32υ23) Ec1(υ12 + υ13υ32) Ec1(υ13 + υ12υ23) 0 0 0
Ec2(υ21 + υ31υ23) Ec2(1 − υ31υ13) Ec2(υ23 + υ21υ13) 0 0 0
Ec3(υ31 + υ21υ32) Ec3(υ32 + υ12υ31) Ec3(1 − υ21υ12) 0 0 0
0 0 0 φGc12 0 0
0 0 0 0 φGc23 0
0 0 0 0 0 φGc31
⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦
(3.8)
where
φ = 1 − υ23υ32 − υ21υ12 − υ31υ13 − υ21υ32υ13 − υ31υ12υ23 (3.9)
CONCRETE 13
The term υij stands for the i-th strain component due to a stress in the j-th direction or simply
Poison’s ratios. The shear moduli are given [8]
Gc12 =Ec1Ec2
Ec1(1 + υ12) + Ec2(1 + υ12)(3.10)
Gc23 =Ec2Ec3
Ec2(1 + υ23) + Ec3(1 + υ32)(3.11)
Gc13 =Ec1Ec3
Ec1(1 + υ13) + Ec3(1 + υ31)(3.12)
where Ec1, Ec2 and Ec3 represent secant elastic moduli of concrete.
For the definition of the secant elastic moduli the reader is referred to [8]. Within a simulation,
the state of stress and strain is transformed into the corresponding principal axis system where
the uniaxial stress-strain relationships are evaluated.
Concrete Constitutive Laws within the 3-D Modified Compression Field Theory
Within the 3-D Modified Compression Field Theory, an uniaxial compressive stress-strain re-
lationship based on the work in [9] is implemented. It reads [8]
σc3 = −fpεc3
εp
n
n − 1 + (εc3/εp)nk(3.13)
where σc3 represents the third principal stress in concrete (compression); εc3 represents the third
principal strain in concrete; fp represents the uniaxial peak compressive stress of concrete; εp
represents the strain in the concrete at fp; n = 0.80; k = 1 for the ascending branch of Eq.
3.13 and k = 0.67 +fp
62for the descending branch of Eq. 3.13.
The uniaxial peak stress fp as well strain εp at peak stress increase if confining stresses exist. In
order to account for confinement effects, a failure criteria is implemented in the 3-D Modified
Compression Field Theory [8]
f = 2.0108J2
f ′c
+ 0.9714
√J2
f ′c
+ 9.1412σc1
f ′c
+ 0.2312I1
f ′c
− 1 = 0 (3.14)
MATERIAL CONSTITUTIVE MODELS 14
where I1 represents the first invariant of the stress tensor; J2 represent the second deviatoric
stress invariant; σc1 represents the first principal stress in concrete and f ′c represents the com-
pressive strength of concrete.
Using the failure criteria Eq. 3.14, the compressive stress σc3,f that causes failure in presence
of stresses σc1, σc2 can be found. The stress σc3,f is used to compute the peak stress factor
Kσ =σc3,f
f ′c
(3.15)
which is used to modify the uniaxial peak strength fp of concrete (Eq. 3.13).
fp = Kσf ′c (3.16)
Due to confinement effects, concrete shows, along with an enhanced uniaxial peak compressive
stress fp, an increasing strain εp at fp. This behavior is captured by mean of the peak strain
factor Kε. For low confining stresses (Kσ < 3):
Kε = 0.2036K4σ − 2.819K3
σ + 13.313K2σ − 24.42Kσ + 13.718
√Kσ + 1 (3.17)
and for larger confining stresses:
Kε = 5Kσ − 4 (3.18)
The strain εp in Eq. 3.13 is then modified:
εp = ε0
{Kσ
(1 − σc3
σc3,f
)+ Kε(
σc3
σc3,f
)}(3.19)
where σc3 represents the current stress in the third principal direction and with the peak strain
under uniaxial compression and the strain at peak stress under uniaxial compression is
ε0 =n
n − 1
f ′c
Ec(3.20)
where n = 0.80 and Ec represents the concrete modulus of elasticity.
CONCRETE 15
3.2.3 Concrete in Tension
Cracks in concrete can be modeled using discrete or smeared cracks. Within the discrete crack
approach concrete cracking is modeled by means of a displacement discontinuity at the in-
terface between concrete elements. The smeared crack concept takes a different approach in
which the cracked material is assumed to be a continuum and the effect of cracking is described
by means of an appropriate tensile stress-strain relationship. The crack is therefore smeared out
over the effected elements. A smeared crack model will be used in this study to model cracking
of concrete.
Within the original smeared crack concept implemented in program DIANA 9.2, the total strain
of a cracked concrete solid ε is decomposed into a part εcr of the crack and a part εco of the
solid material. This allows for implementing constitutive laws incorporating crack dilatancy
which is essentially a coupling between the crack normal direction n and tangential directions
s, t [6]. However, using the total strain based material model for concrete in compression it
is conceptually not possible to combine the approach of decomposed strain in tension with
the total strain approach of concrete in compression. Therefore, a total strain based smeared
crack model [4], available in program DIANA 9.2, is used to model concrete in tension. The
tensile strength of concrete fct is deployed as conditional detection. Once a crack is initiated,
two distinctive approaches as how to handle the crack subsequently exist: the single rotating
crack approach and the single fixed crack approach. In program DIANA 9.2, both approaches
are implemented within the context of total strain based concrete constitutive modeling. In the
single fixed crack approach, the orientation of the crack remains unaltered during the entire
computation process. In the total strain based material models, the stress-strain relationships
of concrete are evaluated in principal stress space. Thus, a disadvantage arises when using
the total strain single fixed crack model. At incipient of cracking, the element principal axes
of strain are replaced by crack directions. Subsequently, the uniaxial constitutive stress-strain
relationships are evaluated in a coordinate system constrained to the crack direction. This yield
a misalignment of the principal directions of stresses and the principal directions of strains in
subsequent load steps. This can lead to spurious high tensile stresses that might exceed the
tensile strength fct multiple times [6]. The misalignment of principal strain and crack direction
can be avoided if a rotating crack model is used. The orientation of the crack rotates with the
axes of principal strain. In conjunction with a tensile cut-off criteria, tensile stresses which
exceed the tensile strength of concrete are then avoided. Thus, the rotating crack model is used
in this study.
MATERIAL CONSTITUTIVE MODELS 16
The post cracking response of reinforced concrete differs from that of plain concrete. While
plain concrete is a brittle material which exhibits softening behavior upon crack initiation if
subjected to uniaxial tensile stress, concrete in RC structures continues to carry tensile stress
between the cracks due to the transfer of forces from the tensile reinforcement to the concrete
through bond. In order to include this ”tension stiffening” effect a material model proposed in
[10] is used. The affiliated material parameters and tensile stress-strain relationship is presented
in section 3.2.4.
3.2.4 Concrete Material Parameters
Concrete Compressive Stress-Strain Relationships
In the side support, center column and the RC beam cover, the compressive relationship pre-
sented in Eq. 3.13 is adopted. The compressive cylinder strength f ′c , listed in table 3.2, has
been determined in compression tests [1]. The modulus of elasticity for concrete is estimated
using
Ec = 57, 000√
f ′c [psi] (3.21)
The strain at peak stress under uniaxial compression ε0 is determined using Eq. 3.20. The
resulting uniaxial stress-strain relationship is shown in Fig. 3.4.
Table 3.2: Material parameters and characteristic values of uniaxial compressive stress-strain
relationship of concrete
Region Ec [ksi] ν f′c [ksi] ε0 εu
Core Concrete 4400 0.15 6.6 0.0022 0.02
Rest of Specimen 4400 0.15 6.0 0.002 0.006
The total strain based material model, as it is implemented in program DIANA 9.2, proved to
be non suitable for the core elements. The approach to modify the peak compressive stress fp if
confining stresses exist, taken in the 3-D Modified Compression Field Theory (Section 3.2.2),
has been completely adopted. Yet, the strain at peak compressive stress εp is modified using
a different approach. The peak strain factor Kε (Eq. 3.17 and 3.18) is set to be equal to peak
stress factor Kσ (Eq. 3.15). However, setting Kε = Kσ is suitable for unconfined concrete
or the cover concrete of the RC beam [8]. For confined concrete, the ratio of strain at peak
compressive stress εp to strain at peak stress under uniaxial compression ε0 increases much
CONCRETE 17
Figure 3.4: Uniaxial compressive stress-strain relationship for concrete cover, side support and
center column
faster than the ratio of peak stress fp to compressive strength of concrete f ′c [8]. Furthermore,
for the enhanced ductility of confined concrete is accounted for using
σc3 = −fp
(1 − (1 − r)
εc3 − εp
εu − εp
)(3.22)
where r represents a factor which models the residual strength and εu denotes the ultimate strain
of concrete
εu =
(fp
fcc
)γ
εp (3.23)
and γ = 3.0.
The ultimate strain εu, predicted using Eq. 3.23, does not lead to sufficient ductility of the
concrete core in the presence of the amount of confinement as is provided by the transverse re-
inforcement of the RC beam. This can be shown by comparing the existing model in program
DIANA 9.2 with Manders model [11], summarized below.
MATERIAL CONSTITUTIVE MODELS 18
The uniaxial compressive stress strain relationship developed in [11] is
σc3 =f ′
ccxr
r − 1 + xr(3.24)
with
x =εc3
εcc
(3.25)
and
εcc = εc
[1 + 5
(f ′
cc
f ′c0
− 1
)](3.26)
where f ′c0 represents the compressive strength of unconfined concrete; εc0 represents the corre-
sponding concrete strain (taken to 0.002); f ′cc represents the compressive strength of confined
concrete and εcc represents the corresponding concrete strain.
In order to estimate the compressive strength f ′cc of confined concrete, the five parameter
William-Warnke failure surface is used [11].
f ′cc = f ′
c0
(− 1.2456 + 2.254
√1 +
7.94f ′l
f ′c0
− 2f ′
l
f ′c0
)(3.27)
where f ′l represents the effective lateral confining stress.
The effective lateral confining stress is a function of the yield stress fyh, of the transverse
reinforcement, the confinement effectiveness coefficient ke and the ratio ρ of the volume of
transverse confining steel to the volume of the RC beam core.
f ′l = keρfyh (3.28)
The confinement effectiveness coefficient is
ke =Ae
Acc
(3.29)
with
Ae =
(1 −
n∑i=1
(w′i)
2
6bcdc
)(1 − s′
2bc
)(1 − s′
2dc
)(3.30)
CONCRETE 19
Acc = Ac(1 − ρcc) (3.31)
where Ae represents the area [in2] of the effectively confined concrete core; Acc represents the
area [in2] of the core of the RC beam enclosed by the center lines of the perimeter hoop and ρcc
represents the ratio of the area of the longitudinal reinforcement to the area of the core of the
RC beam section.
In Fig. 3.5, the cross section of the RC beam at the face of the center column is depicted. The
height of the compressive zone at peak compressive stress is estimated to a third of the section
height. The strain at the top of the core section is assumed to be at εco = 0.002.
Figure 3.5: Effectively confined concrete core
The resulting uniaxial compressive stress-strain relation of the concrete core and the trilinear
approximation, which is used in the finite element model, are shown in Fig. 3.6. The material
parameters for the concrete core and the characteristic values of the uniaxial compressive stress-
strain relationship are listed in Table 3.2.
Concrete Tensile Stress-Strain Relationship
A rotating crack model, outlined in section 3.2.3, is adopted in order to incorporate cracking
of the concrete in the RC beam. In order to account for tension stiffening effects of reinforced
concrete in tension, which contributes to the overall flexural stiffness of the beam, a piecewise
linear tensile stress-strain relationship developed in [10] is adopted. The model is depicted in
Fig. 3.7 and the characteristic values are listed in Table 3.3. The tensile strength used as tension
cut-off condition is
f ′ct = 6.45 ·
√f ′
c [psi] (3.32)
MATERIAL CONSTITUTIVE MODELS 20
Figure 3.6: Uniaxial compressive stress-strain relationship for concrete core [11] and the tri-
linear approximation
Figure 3.7: Uniaxial tensile stress-strain relationship of reinforced concrete [10]
Table 3.3: Parameters of tensile stress-strain relationship for reinforced concrete
Pt Rt St Ft εcr fct [Ksi ]
0.8 0.45 4 10 1.4e-4 0.5
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 21
3.3 Interface between Concrete and Steel Reinforcement
In order to capture the contribution of bond to the flexibility and deformation capacity of the
beam, the interface between concrete and steel reinforcement elements is explicitly modeled
by means of three-dimensional interface elements.
3.3.1 Mechanics of Interaction between Steel Reinforcement and Con-
crete
In RC elements, at the interface between the rebars and concrete, three mechanisms are acti-
vated if the rebars are subjected to tensile or compressive stresses: chemical adhesion, mechan-
ical interaction between the rebar ribs and the concrete keys as well as friction [12]. Fig. 3.8
schematically shows bond stress-slip relationships [13].
For bond stress values τ ≤ τI , the concrete surrounding the rebar remains in the elastic stage
and chemical adhesion governs the transfer of the rebar stress into the surrounding concrete.
The slip between the concrete and the rebar remains small [13].
Splitting
Failure
Splitting induced
Pull-Out Failure
Pull-OutFailure
Confinement
0.5 f’ c
�I I
III
IV a
IV b
Av. BondStress
Slippage s
Curve 1 Curve 2 Curve 3�I
II
Figure 3.8: Qualitative local bond stress-slip relationships [13]
For bond stress values τ > τI , the capacity of the chemical adhesion to transfer the rebar stress
into the concrete is exceeded and bearing stresses in front of the rebar ribs are generated (me-
chanical interaction). Transverse cracks, shown in Fig. 3.9, arise in the surrounding concrete
MATERIAL CONSTITUTIVE MODELS 22
causing the rebar to slip. As the bond stress τ increases, the concrete subjected to bearing
stresses crushes, shown in Fig. 3.9, and the bond stiffness decreases (Fig. 3.8, Stage II).
T e n s io n
C o m p re s s io n
R e b a r
T ra n s v e rs eM ic ro c ra c k s
c ru s h e dC o n c re te
Figure 3.9: Transverse cracks originated at the tips of the steel reinforcement ribs and crushed
concrete in front of the ribs of the rebar [12]
The outward component of the bond stress is balanced by ring tensile stresses [14], shown in
Fig. 3.10. If the tensile stresses become large enough, longitudinal cracks develop along the
rebar. If no or an insufficient amount of transverse reinforcement is provided, the longitudinal
cracks reach the outer concrete surface and bond failure occurs. (Fig. 3.8, Stage III, Curve 1).
As the amount of transverse reinforcement increases, the spreading of the longitudinal cracks
is increasingly restricted as shown in Fig 3.11. The longitudinal cracks do not reach the outer
concrete surface and as a result, larger bond stresses can develop [13] (Fig. 3.8, Stage IVa,
Curve 2). The development of longitudinal cracks remains within an small area around the
rebar section if the confinement action, either provided by transverse reinforcement or trans-
verse pressure, further increases. More transverse cracks in the surrounding concrete arise and
the crushing of the concrete in front of the rebar ribs continues [12]. This bond mechanism
provides the largest bond strength (Fig. 3.8, stage IVb, Curve 3). The magnitude of bond stress
in stage IVb can be as high as 0.5f ′c.
At maximum bond stress τmax, the concrete keys between the rebar ribs start to shear off. With
increasing slip, an increasingly larger part of the concrete keys are sheared off and the bond
stress decreases until only the frictional part of the bond resistance τresidual remains when the
concrete surface is smoothed out [12] (Fig. 3.8, Curve 2 and 3).
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 23
Figure 3.10: Radial component of
bond stress balanced by tensile ring
stresses in the surrounding concrete
[14]
Figure 3.11: Radial cracks induced by
tensile ring stresses and transverse re-
inforcement action [13]
3.3.2 Review of Existing Bond Stress-Slip Relationships
In [12], pull-out tests have been conducted using a specimen aimed to simulate conditions
within a beam-column joint. The specimen consisted of a single deformed rebar embedded
in a concrete block over a length of le = 5db. The bond stress along the embedment length
was assumed to be evenly distributed due to the short anchorage length. For the majority of
the tests, the concrete block was reinforced with longitudinal (perpendicular to the axis of the
anchored rebar) and transverse (parallel to the axis of the anchored rebar) steel reinforcement.
For the remaining tests the concrete block was not reinforced in order to obtain reference data
for unconfined concrete. One protruding end of the anchored rebar was subjected to load under
displacement control while the slip was measured as the movement of the unloaded end of the
rebar with respect to the concrete anchorage block. The load applied to the rebar was reacted
as a compressive force at the face of the concrete block. Bond stress-slip relationships were
deduced by taking applied forces at given slip value and converting them into bond stress using
Eq. 3.33.
τ =F
π · db · le (3.33)
where F represents the applied force; db represents the rebar diameter and le represents the
embedment length.
MATERIAL CONSTITUTIVE MODELS 24
The test program consisted of several series each designed to investigate a single varying pa-
rameter that influences bond characteristic while all other specimen parameter were kept con-
stant. The longitudinal reinforcement varied from #2 to #8 rebars and the transverse reinforce-
ment varied from #2 to #4 stirrups. Furthermore, the diameter of the anchored rebar varied
from 0.5 in to 1.0 in in order to obtain information with regard to the influence of the rebar di-
ameter on the bond characteristics. Additionally, a test series was conducted where the specific
rib area varied from fr= 0.066 to fr= 0.12. The specific rib area is
fr =Ar
π · db · sr
(3.34)
where Ar represents the area of the projection of a single rib on the cross-section of the rebar;
sr represents the rib spacing and dbrepresents the rebar diameter.
A test series with concrete compressive strength varying from f ′c=4350 psi to f ′
c=7975 psi was
also conducted. In all tests, a splitting crack developed in the plane of the longitudinal axis of
the anchored rebar and the failure of bond was observed if the concrete was not confined by
reinforcement. The concrete between the ribs of the rebar was intact. In case of confined con-
crete, the growth of the splitting crack was controlled by the longitudinal reinforcement. The
load could be increased further and failure of bond was caused by pulling out of the anchored
rebar. It is reported that the concrete between the rebar ribs was completely sheared off and
almost pulverized. As a result of these tests, Eq. 3.35 was proposed to represent an average of
experimentally obtained bond stress-slip curves for confined concrete.
τ = τ1 ·(
s
s1
)α
; s ≤ s1 (3.35)
where τ represents the bond stress; s represents the slip of the rebar; τ1 = 13.5N/mm2;
s1 = 1 mm and α = 0.4
According to [12], bond between rebars and concrete scatters. The values for τ1 and α varied
between τ1 ≈ 11.5N/mm2 to τ1 ≈ 15.5 N/mm2 and α ≈ 0.33 to α ≈ 0.45. It is reported that
the influence of the the rebar diameter on bond characteristics was rather small. The values of
τmax increased approximately proportional to√
f′c and the corresponding slip values decreased
almost proportional to 1√f′c
. Varying the clear spacing between the rebars from a minimum
value of sb = 1db to a maximum value of sb = 4db resulted in an increase of τmax of about
20%. It is furthermore reported that the distance between the rebar ribs sr greatly influences
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 25
the characteristic of bond. The slip value s1 at which τmax was reached increased if sr was
enlarged. Thus, modification factors are proposed which may be applied to Eq. 3.35 if it is
intended to use the bond stress-slip relationship (Eq. 3.35) in specific cases where different
confining reinforcement details, rebar diameter, deformation pattern and concrete compressive
strength prevail. These modification factors are presented and applied later in this chapter.
In [15], three direct tension tests were conducted. The test set up was designed to simulate
conditions prevailing in a region of a beam between flexural tensile cracks. The specimen con-
sisted of a single #8 rebar centrally embedded over a length of 18db in a cross section concrete
block of 5x5x18 in. The rebar was subjected to tensile forces at the projecting ends. A bond-
stress slip relationship was determined using a different technique compared to aforementioned
pull-out test. The rebar was equipped with internal strain gages which were distributed along
the embedment length. Additionally strain gages were embedded in the concrete block at a
distance of 0.25 in from the rebar surface. The concrete strain and the strain of the embedded
rebar were measured and recorded at different locations along the embedment length. Subse-
quently, each strain function was integrated resulting in the displacement of the concrete and
the rebar. The slip was then computed by subtracting the concrete displacements from the re-
bar displacements. The findings of this study show that the bond stress-slip relationship varies
significantly along the embedment length. The maximum bond stress varied from 4√
f ′c at a
distance of 2 in from the loaded end of the rebar to 10√
f ′c at a distance of 6 in from the loaded
end of the rebar. The slip at which τmax was reached varied from 0.0005 in at 2 in to 0.0015 in
at 6 in from the loaded end of the rebar. Using the test results, Eq. 3.36 was proposed in [16]
to represent a best fit to the average of the experimental results.
τ = f ′c · (16.7s − 8260s2 + 1.12 · 106s3) [ksi] (3.36)
where f ′c represents concrete compressive strength and s represents the rebar slip.
In [17], a series of direct tension tests on concrete square prisms was conducted. Each specimen
was axially reinforced with one #8 central rebar. The rebar was instrumented with internal
strain gages. This technique allowed for obtaining the steel stress distribution along the length
of the rebar. The length of the concrete prism was 16 in which yield an embedment length of
16db. Cross sectional dimensions varied from 2x2 in to 6x6 in which led to different ratios of
concrete cover to rebar diameter. The specific rib area fr for all rebars was nearly constant at
fr=0.14 to 0.15. The concrete compressive strength varied from f ′c = 4300-5000 psi. The slip
MATERIAL CONSTITUTIVE MODELS 26
of the rebar was measured at 0.25 in from the rebar surface using a micrometer. Bond stress-slip
relationships were deduced by using Eq. 3.33. The force F in Eq. 3.33 was derived as force
transferred into the concrete using information provided by the stress distribution along the
rebar. It is reported that in all tests no crushing of concrete in front of the rebar lugs occurred.
Therefore, it is concluded that the observed slip is due to the internal cracking of the first layer
of the concrete surrounding the rebars (see Fig. 3.15) and due to bending and/or cracking of the
concrete keys near the rebar ribs. Furthermore, it was observed that the concrete compressive
strength had an insignificant effect on measured slip values. Yet, the maximum bond stress was
observed to be proportional to√
fc. It is also reported that with increasing cover thickness, the
slip at which maximum bond stress was reached decreased owing to the enhanced restraining
capacity of the concrete mass. As a result of this study, Eq. 3.37 was proposed to represent
a best fit to experimentally obtained data which were normalized to a concrete compressive
strength of f ′c = 5000 psi.
τ = 1.95 · 106s − 2.35 · 109s2 + 1.39 · 1012s3 − 0.33 · 1015s4 [psi] (3.37)
where f ′c represents concrete compressive strength and s represents the slip of the rebar.
In Fig. 3.12 the bond stress-slip relationships are shown for two slip ranges. As can be seen,
the bond stress-slip relationships differ considerably. For very small values of up to a slip of
about 0.0001 in the bond stiffness is rather close. But with increasing slip values the difference
in the tangent modulus is apparent. According to [12], a large scatter in the initial bond stiff-
ness may be caused by inaccuracies in measuring the slip between the rebar and the concrete
correctly. If slip values are deduced from measured concrete and rebar strains, as conducted in
[15], the error even accumulates. Also, the bond stress-slip relationship for rebars embedded
over a length > 5db cannot be considered as constant along the embedment length. It varies
considerably at distinct locations along the rebar. Furthermore, it is reported that the position
of the rebars during casting influences the initial bond stiffness [12]. Rebars cast horizontally
show much smaller initial bond stiffness compared to rebars cast vertically.
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 27
Figure 3.12: Bond stress-slip relationships [12], [15] and [17]
Another potential source of the scatter in initial bond stiffness is the use of different test speci-
mens with different stress conditions prevailing in the concrete surrounding the rebar. While in
pull-out tests the concrete is subjected to longitudinal compression, in direct tension tests the
concrete is subjected to tensile stresses. Experimental findings indicate that one of the main
source of scatter in initial bond stiffness is the use of rebars with different deformation pattern
which lead to different specific rib area fr for identical rebar diameter [12]. Larger specific rib
area yield larger bearing area of the rebar ribs which in turn induces smaller bearing stresses
on the concrete between the rebar ribs at equal rebar stress level. This reduces the strain of
the concrete in the vicinity of the rebar which lead to smaller slip values at the steel-concrete
interface. Additionally, it is reported that for identical specimens maximum bond stress τmax
was reached at slip values of smax= 0.028 in and smax=0.054 in for rebars having a specific rib
area of fr = 0.12 and fr = 0.066, respectively [12]. This corresponds to an increase in initial
bond stiffness of about 50% due to an increase in specific rib area of about 50%. The specific
rib area reported in [12] and [17] varied from fr = 0.065 to fr = 0.14 − 0.15, respectively,
whereas the diameter of the rebar in both test series was the same.
3.3.3 Parameters influencing Bond Performance
Geometry and Deformation Pattern of Steel Reinforcement
As summarized and discussed in the previous section, the rib geometry of the rebar has paramount
importance among the other parameters that effect bond behavior. It has been found that the
maximum bond stress τmax strongly depends on the specific rib area fr. Experimental results
[12] indicate that the slip smax at which τmax is reached or the initial bond stiffness, respectively,
MATERIAL CONSTITUTIVE MODELS 28
increases as the specific rib area increases. Furthermore, according to [12], the rib spacing
sr has a significant effect on the characteristic values of a bond stress-slip relationship. With
increasing values of sr, the slip smax at which maximum bond stress τmax is reached, increases.
Similarly, the slip at which the bond stress-slip relationship levels off to the frictional bond
stress τresidual, increase.Another important factor is the cover thickness. With an increasing
cover thickness, the restraining effect on the rebar is enhanced [17]. The level of stress, present
in the rebar, also effects the bond performance. As long as the rebar remains within the linear
elastic range, the influence of steel reinforcement stress remains small. Yet, experimental tests
show that yielding of the steel reinforcement has a negative effect on bond. It results in a sharp
nonlinear descending branch in the bond stress-slip relationship once yielding of the rebar has
occurred [13].
Concrete Characteristics
The maximum bond stress τmax varies with increasing concrete compressive strength f ′c [12].
It has been experimentally shown that τmax is proportional to√
f ′c [12] and [17]. This is
because bond action results from the localized pressure induced into the concrete in front of
the rebar ribs and the pressure is directly related to the shear component of bond stress [13].
Furthermore, experimental results indicate that slip values corresponding to τmax decrease ap-
proximately proportional to 1/√
f ′c which results in an increasing bond stiffness [12].
Confinement Effects
Transverse compressive stresses favor bond action independent on whether they result from
active or passive confinement. Active confinement, resulting from a direct support or from a
continuity of a column in a beam-column joint, is more efficient to prevent splitting failure
since it does not depend on the bond stress itself. On the contrary, passive confinement devel-
oped by concrete cover and transverse reinforcement is less effective since it has its origin in
the dilatancy of concrete cracks and the development of cracks is related to bond stress [13].
Passive confinement controls the spreading of the longitudinal splitting cracks and prevents
pure splitting failure, as reported in [12]. Experimental tests show that if sufficient passive
confinement is provided or the stirrup confinement index Ω is within a certain range, maximum
bond strength increases to a certain extent [13].
Additional factors influencing bond characteristic are the clear rebar spacing sb as well as the
rebar diameter. In [12], a decreasing maximum bond stress is reported if the clear rebar spacing
sb falls below 4db. A slight decrease of the maximum bond stress of about 10% for increasing
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 29
rebar diameter and equal values of relative rib area fr was observed in [12]. Furthermore, rebar
corrosion, rusting and the loading also affect bond-slip [13].
3.3.4 Implemented Bond Stress-Slip Relationships
The constitutive law, available in program DIANA 9.2, which governs the behavior of the 3-
dimensional interface elements is [8]
(τn
τt
)=
⎛⎝kn 0
0∂τ(st)
∂st
⎞⎠ ·
(Δsn
Δst
)(3.38)
where τn, τt represent the normal and the tangential component of the interface traction or the
bond stress; τ(st) represents the bond stress-slip relationship provided as input; kn represents
the normal stiffness of the interface between concrete and rebar and sn, st represent the normal
and the shear component of the interface relative displacement or the bond slip.
As can be seen, the shear and normal components are decoupled. Thus, the interaction between
the shear component of bond and the lateral contraction and extension of the rebars due to Poi-
son’s effect cannot be modeled. Furthermore, no communication protocol between interface
elements and concrete elements or steel reinforcement elements is implemented in program
DIANA 9.2. Thus, bond deterioration due to reinforcement yielding and concrete crushing
cannot be considered.
In the finite element model of the RC beam, two distinctive regions with regard to bond con-
ditions are defined. Region A (Fig. 3.13) is defined well as confined. Given the existence of
longitudinal and transverse reinforcement, it is assumed that the bond characteristic in region
A is of a pull-out failure type [13]. The bond stress-slip relationship for the interface elements
around the rebars subjected to compressive stress and for the interface elements around the
rebars subjected to tensile stress are assumed to be equal which complies with experimental
findings [12]. Table 3.4 provides the values of τmax, τresidual and of s1, s2 and s3 in region A,
which are adopted according to the test data obtained in [12]. Note that the pattern of the bond
stress-slip relationship in both regions is the same. The pattern was proposed in [12] and is
shown in Fig. 3.14 along with the designation of the characteristic points of the curve. The
concrete compressive strength f ′c , the clear bar spacing sb and the rebar diameter db differ from
those prevailing in the pull-out tests [12]. Furthermore, the deformation pattern of the rebars
used in [1] is different from that of the rebars used in [12].
MATERIAL CONSTITUTIVE MODELS 30
Region B
Region ARegion A S 3S 2S 1 Slip s
m ax
BondStress
res idua l
τ
τ
Figure 3.13: Defined regions of bond
in the finite element model
Figure 3.14: Pattern of the bond
stress-slip relationship in region A
and B
Table 3.4: Initial characteristic values of the bond stress-slip relationships in region A and B
Region τmax [psi] τresidual [psi] s1 [in] s2 [in] s3 [in]
A 1640 410 0.035 0.07 0.4
B 800 200 0.01 0.02 0.1
Thus, the initial values of the characteristic points of the bond stress-slip relationship in region
A, given in Table 3.4, are subsequently modified according to suggestions made in [12] in order
to account for differences in aforementioned parameters.
Region B within the finite element model of the RC beam is defined. As discussed earlier, the
values of τmax deduced from direct tension tests are considerable smaller than the values for
τmax obtained in pull-out test in the confined concrete blocks. A bond model presented in [14]
is used to approximate the initial value of τmax in region B. In the model, a cracked concrete
sleeve around the rebar is restraint by an outer solid sleeve, subjected to tensile hoop stresses,
such that the rebar does not slide out of the concrete [14] (Fig. 3.15). The initial value of τmax
is estimated using Eq. 3.39. The value of τmax represents the bond stress at the incident of
bond failure of a embedded rebar once the thickness of the inner sleeve reaches a critical value,
namely 1.664 db (Fig. 3.15). The transverse reinforcement of the tested RC beam comprised
with a diameter of 0.135 in placed at 2.7 in. This is considerably less than in region A. Thus, the
restraining effect on the concrete surrounding the rebars is not as pronounced as in region A.
However, it provides to some extent additional restraining effects yielding a descending branch
of the bond stress-slip relationship and an increased maximum bond stress τmax.
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 31
τmax = fct ·cy +
db
21.664 · db
= 800 psi (3.39)
where
cy = 0.8175 in concrete cover of the RC beam
db = 0.375 in rebar diameter
fct = 500 psi concrete tensile strength
Figure 3.15: Hydraulic pressure anal-
ogy taken from [14]
The initial value of τmax = 800 psi compares reasonable to experimental data presented in sec-
tion 3.3.2 (Fig. 3.12), given the scatter in maximum bond stress. With regard to the initial
values of slip in region B, data given in [15] and [17] vary significantly as presented earlier
and do not seem to be reliable. Therefore, the initial value of s1 in region B is adopted based
on a pull-out test on concrete specimen without confining reinforcement [12]. Values for the
slip s2 and s3 are also suggested. However, these values are based on data which were obtained
in pull-out tests showing splitting failure of the specimen yielding a sharp descending branch
of the bond stress-slip relationship (in Fig. 3.14 s2 = s1). Yet, due to the confinement effect,
exerted by the transverse reinforcement of the RC beam, the pattern of the bond stress-slip
relationship in region B is equal to that of region A. Thus, the ratio of s2 to s1 and of s3 to s1 in
region A, which computes to 2 and 11.5, respectively, is used in order to obtain initial values
for s2 and s3 in region B. Following the same approach, the ratio of τmax to τresidual in region A,
which computes to 4, is used to obtain an initial value of τresidual in region B.
Modification of Initial Characteristic Values due to Concrete Compressive Strength
The compressive strength of the concrete of the RC beam is f ′c = 6 ksi. The initial value for
τmax in region A is based on pull-out tests in specimens having a concrete compressive strength
of f ′c = 4.35 ksi. In order to convert the initial values to account for the concrete strength of
the RC beam, τmax and τresidual in region A are increased by a factor of about√
64.35
= 1.17
MATERIAL CONSTITUTIVE MODELS 32
[12]. In region B, τmax is not modified since Eq. 3.39 was developed based on pull-out tests in
specimens having also a compressive strength of f’c = 6 ksi. The initial value for s1 in region
A and B is reduced by a factor of 11.17
based on experimental findings in [12].
Modification of Initial Characteristic Values due to Confining Reinforcement
In [12], the ratio
∑Asv∑As
(3.40)
where Asv represents the area of the confining reinforcement and As represents the area of the
anchored RC beam reinforcement, varied between 0.0 to 4.0. Eq. 3.40 gives a value of about
1.6 for the bottom reinforcement of the RC beam and a value of about 4.0 for the top reinforce-
ment. These values compare favorable to the values in [12]. Thus, the initial characteristic
values of τmax and τresidual in region A are not modified.
The stirrup confinement index Ω is [13]
Ω =Ast
A∗ (3.41)
where Ast represents the area of the transverse reinforcement; A∗ = n · db · Δz represents the
area of the rebars in the splitting plane; n represents the number of rebars; db represents the
rebar diameter and Δz = 2.7 in represent the distance at which the transverse reinforcement is
placed along the RC beam (Fig. 2.5, Page 3).
For the RC beam, Ω is 0.006 and 0.015 for five #3 rebars (top of section) and two #3 rebars
(bottom of section). Experimental results indicate that for Ω between 0.006 to 0.014 a maxi-
mum bond stress τmax of 0.12 f ′c to 0.18f ′
c can be expected [13]. This yield a maximum bond
stress of τmax ≈ 720−1080 psi considering that f ′c = 6000 psi. The initially estimated value of
τmax = 800 psi represents a lower bound of maximum bond stress [14]. Thus, the initial char-
acteristic value of τmax in region B is increased by 200 psi in order to account for confinement
effects exerted by the transverse reinforcement of the RC beam.
Modification of Initial Characteristic Values due to Deformation Pattern
The specific rib area fR of the #3 rebars used in [1] is 0.1. fR of the rebars used in [12] is 0.11.
Thus, no modification factor is applied to τmax in region A.
INTERFACE BETWEEN CONCRETE AND STEEL REINFORCEMENT 33
The rib spacing sr of the rebars used in [1] is sr = 0.2 in. Yet, the rib spacing of the rebars
used in [12] is sr = 0.4 in. Thus, the initial characteristic values for s1, s2 and s3 in region A
and B are modified by a factor of 0.2 in0.4 in = 0.5, as suggested in [12].
Modification of Initial Characteristic Values due to Rebar Spacing
In the RC beam, the clear rebar spacing sb is 0.9 in for the top rebars and 4.27 in for the bottom
rebars. The ratio
sb
db
(3.42)
where db represents the rebar diameter,is computed to be 2.4 and 11.38 for the top and the
bottom rebars, respectively. Experimental findings suggest a reduction of τmax and τresidual by
10% for values of 2.4 [12]. This reduction is applied to the initial values in region A and B.
The final values for the characteristic points of the bond stress-slip relationships, which are im-
plemented in the finite element model, are listed in Table 3.5 and Fig. 3.16 shows the affiliated
bond stress-slip relationships.
Table 3.5: Characteristic values of interface bond stress-slip relationships of region A and B as
implemented in the finite element model
Region τmax [psi] τresidual [psi] s1 [in] s2 [in] s3 [in]
A 1740 500 0.016 0.035 0.2
B 1000 250 0.004 0.008 0.05
MATERIAL CONSTITUTIVE MODELS 34
Figure 3.16: Bond stress-slip relationship of interface elements in region A and B as imple-
mented in the finite element model
Chapter 4 Analytical ResultsIn this chapter, the variation of stress, strain, bond stress, and slippage along the beam are pre-
sented. Fig. 4.2 shows the locations where the results for concrete, steel and interface elements
were obtained. Contour plots of two longitudinal profiles are also presented. Longitudinal
profile 1 includes core and cover concrete whereas longitudinal profile 2 includes only cover
concrete, as depicted in Fig. 4.1. The results in this chapter are presented for different load
steps (LS). The vertical displacement associated with each load step is listed in Table 4.1.
Figure 4.1: Designation of longitudi-
nal views used for contour plots
Figure 4.2: Locations of concrete C,
steel S and interface I elements
Table 4.1: Analysis load steps and associated vertical displacement at the face of center column
Load Step (LS) 1 2 3 4 5 6 7 8 9
Vertical Displacement [in] 0.15 0.3 0.48 0.76 1.0 1.2 1.45 1.68 1.9
Load Step (LS) 10 11 12 13 14 15 16 17
Vertical Displacement [in] 2.13 2.35 2.58 2.8 3.21 3.62 4.44 5.26
ANALYTICAL RESULTS 36
4.1 Analytical Results at Material Level
In this section, analytical results along the beam are presented. Note that the center line of the
two-span RC beam (where the vertical displacement is applied) is located at zero distance (see
Fig. 2.5). The face of the side support is at a distance of 82 in.
Concrete in Tension
Behavior of concrete in tension is modeled using rotating cracks (Sec. 3.2.3). Thus, at every
location within the beam, tensile stresses are expected to be less than the tensile strength of
fct = 500 psi. In Fig. 4.3, the variation of concrete longitudinal stress σc,zz along the beam
at location C3 is shown for three load steps LS1, LS2 and LS3 (see Table 4.1). As can be
seen, longitudinal tensile stresses σc,zz do not exceed the tensile strength fct = 500 psi at
any location. The distance between location of peak tensile stress of 500 psi and the face of
the center column in Fig. 4.3 can be considered as the length over which the concrete tensile
strain has exceeded the crack strain of εcr = 0.00014 and the concrete elements go through
the descending branch of uniaxial tensile stress-strain relationship (see Fig. 3.7). The tensile
stress equal to zero indicates that concrete tensile strain has exceeded ultimate tensile strain
εu,cr = 0.0014 (see Fig. 3.7). Within the distance between the peak tensile stress of 500 psi
and inflection point (zero bending moment), concrete elements are in the pre-cracking elastic
stage. As can be seen in Fig. 4.4, for larger deformations, tensile stresses no longer reach the
tensile strength of fct = 500 psi. This indicates that all concrete elements between the face of
the center column and inflection point and top bar cut-off location at C3 have passed the crack
strain of εcr = 0.00014. Three contour plots of concrete first principal stress σc,1, shown in
Fig. 4.5, also confirm that stress values have not exceeded the tensile strength of fct = 500 psi
throughout the analysis.
ANALYTICAL RESULTS AT MATERIAL LEVEL 37
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−3500
−3000
−2500
−2000
−1500
−1000
−500
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TR
ES
S [p
si]
CONCRETE CORE BOT MIDDLE STRESS OVER LS
LS1LS2LS3
Faceof
CenterColumn
Faceof
SideSupport
Top Bar Cut−OffLocation
Top Bar Cut−OffLocation
Figure 4.3: Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at 0.15,
0.3 and 0.48 in of vertical displacement
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−6,000
−5,000
−4,000
−3,000
−2,000
−1,000
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TR
ES
S [p
si]
CONCRETE CORE BOT MIDDLE STRESS OVER LS
LS7LS8LS9
Faceof
CenterColumn
Faceof
SideSupport
Top Bar Cut−OffLocation
Top Bar Cut−OffLocation
Figure 4.4: Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at 1.45,
1.68 and 1.9 in of vertical displacement
ANALYTICAL RESULTS 38
-.2E4-.1E4-5000250500
-.2E4-.1E4-5000250500
-.2E4-.1E4-5000250500
Figure 4.5: Longitudinal profile 1 of first principal stress σc,1 [psi] at three vertical displace-
ments: 0.76 (Top); 1.0 (Middle) and 1.2 in (Bottom)
Concrete in Compression
Figs. 4.6 and 4.7 show the variation of concrete longitudinal stress σc,zz along the beam at cross
sectional locations C1 and C2 (see Fig. 4.2), respectively. Maximum concrete compressive
stress of σc,zz = -5227 psi (Fig. 4.6) and σc,zz = -5789 psi (Fig. 4.7) are calculated to occur
at 1.9 in of vertical displacement for location C1 and C2, respectively. For concrete elements
at cross sectional location C3 and C4, maximum concrete compressive stress in longitudinal
direction of σc,zz = -5911 psi (Fig. 4.8) and σc,zz = -6278 psi (Fig. 4.9) are calculated at a
vertical displacement of 1.2 in, respectively. The peak compressive strengths of f ′c= 6000 psi
for cover concrete and f ′c=6600 psi for core concrete have not been exceeded throughout the
analysis. With regards to third principal stress, maximum values of σc,3 = 6000 psi and σc,3 =
6600 psi for cover concrete and core concrete, respectively, are calculated (Figs. 4.10 and
4.11). This is consistent with the Modified Compression Field Theory where the state of stress
is evaluated in principal stress space. Values for the third principal stress larger than the strength
of uniaxial compressive stress-strain relationship are not expected to occur.
ANALYTICAL RESULTS AT MATERIAL LEVEL 39
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90-5500
-5000
-4500
-4000
-3500
-3000
-2500
-2000
-1500
-1000
-500
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TRE
SS
[psi
]
CONCRETE COVER TOP MIDDLE STRESS OVER LS
LS7LS8LS9
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut-Off
Location
BarCut-Off
Location
Figure 4.6: Variation of concrete longitudinal stress σc,zz at cross sectional location C1 at 1.45,
1.68 and 1.9 in of vertical displacement
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90-6000
-5500
-5000
-4500
-4000
-3500
-3000
-2500
-2000
-1500
-1000
-500
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TRE
SS
[psi
]
CONCRETE CORE TOP MIDDLE STRESS OVER LS
LS7LS8LS9
Faceof
CenterColumn
Faceof
SideSupport
Top BarCut-Off
Location
Top BarCut-Off
Location
Figure 4.7: Variation of concrete longitudinal stress σc,zz at cross sectional location C2 at 1.45,
1.68 and 1.9 in of vertical displacement
ANALYTICAL RESULTS 40
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−6000
−5500
−5000
−4500
−4000
−3500
−3000
−2500
−2000
−1500
−1000
−500
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TR
ES
S [p
si]
CONCRETE COVER BOT MIDDLE STRESS OVER LS
LS4LS5LS6
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
SideSupportFace
ofCenter
Column
Figure 4.8: Variation of concrete longitudinal stress σc,zz at cross sectional location C4 at 0.76,
1.0 and 1.2 in of vertical displacement
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−6500
−6000
−5500
−5000
−4500
−4000
−3500
−3000
−2500
−2000
−1500
−1000
−500
0
500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E S
TR
ES
S [p
si]
CONCRETE CORE BOT MIDDLE STRESS OVER LS
LS4LS5LS6
Faceof
SideSupport
Faceof
CenterColumn
Top Bar Cut−OffLocation
Top Bar Cut−OffLocation
Figure 4.9: Variation of concrete longitudinal stress σc,zz at cross sectional location C3 at 0.76,
1.0 and 1.2 in of vertical displacement
ANALYTICAL RESULTS AT MATERIAL LEVEL 41
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
Figure 4.10: Longitudinal profile 1 of third principal stress σc,3 at 1.2, 1.45, 1.68 and 1.9 in
(plots from top to bottom) of vertical displacement (the face of the side support is located on
the left, the face of the center column is located on the right)
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
-.66E4-.6E4-.4E4-.2E40500
Figure 4.11: Longitudinal profile 2 of third principal stress σc,3 at 1.2, 1.45, 1.68 and 1.9 in
(plots from top to bottom) of vertical displacement (the face of the side support is located on
the right, the face of the center column is located on the left)
ANALYTICAL RESULTS 42
Steel Reinforcement
Figs. 4.12 and 4.13 show the variation of longitudinal strain εs,zz and stress σs,zz , respectively,
for steel reinforcement elements at cross sectional location S1 (see Fig. 4.2). At the vicinity
of the side support, a maximum tensile strain of εs,zz = 0.0341 is calculated at a vertical dis-
placement of 5.26 in (LS17). At the vicinity of the center column, a maximum compressive
strain of εs,zz = -0.0355 is computed at the same level of vertical displacement. Maximum
steel reinforcement stress values in longitudinal direction of σs,zz = 92300 psi in tension and
σs,zz = -94500 psi in compression are calculated at a vertical displacement of 5.26 in.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.04
−0.03
−0.02
−0.01
0
0.01
0.02
0.03
0.04
DISTANCE ALONG BEAM [in]
AV
ER
AG
E E
LEM
EN
T S
TR
AIN
S1 Strain Variation
LS16
LS17
Top BarCut−OffLocation
Faceof
SideSupport
Top BarCut−OffLocation
Faceof
CenterColumn
Figure 4.12: Variation of steel longitudinal strain at cross sectional location S1 at 4.44 and 5.26
in of vertical displacement (horizontal dashed lines represent yield strains)
Fig. 4.14 shows the variation of longitudinal strain εs,zz for steel reinforcement elements at
cross sectional location S2 (see Fig. 4.2). At a vertical displacement of 5.26 in, a maximum
tensile strain of εs,zz = 0.13 is computed at the vicinity of the center column and a maximum
compressive strain of εs,zz = -0.0586 at the vicinity of the side support. Fig. 4.15 shows the
tensile stress-strain relationships of the first three steel reinforcement elements from the face of
the center column. As can be seen, the bar fracture occurs in the second element at a vertical
displacement of 5.26 in.
ANALYTICAL RESULTS AT MATERIAL LEVEL 43
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−95,000
−80,000
−60,000
−40,000
−20,000
0
20,000
40,000
60,000
80,000
95,000
DISTANCE ALONG BEAM [in]
AV
ER
AG
E E
LEM
EN
T S
TR
ES
S [p
si]
S1 Stress Variation
LS16LS17
Faceof
SideSupport
Top BarCut−OffLocation
Faceof
CenterColumn
Top BarCut−OffLocation
Figure 4.13: Variation of steel longitudinal stress at cross sectional location S1 at 4.44 in and
5.26 in of vertical displacement (horizontal dashed lines represent yield stresses)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.06
−0.04
−0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
DISTANCE ALONG BEAM [in]
AV
ER
AG
E E
LEM
EN
T S
TR
AIN
S2 Longitudinal Strain Variation
LS16LS17
Top BarCut−OffLocation
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut−OffLocation
Figure 4.14: Variation of steel longitudinal strain at cross sectional location S2 at 4.44 and 5.26
in of vertical displacement (horizontal dashed lines represent yield strains)
ANALYTICAL RESULTS 44
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.140
1
2
3
4
5
6
7
8
9
10
11x 10
4
STRAIN
ST
RE
SS
[psi
]
STEEL STRESS−STRAIN
1st Element2nd Element3rd ElementInput Uniaxial Stress−Strain
Figure 4.15: Tensile stress-strain of the first three steel reinforcement elements from the face
of the center column at cross sectional location S2. (Uniaxial stress-strain relationship of steel
reinforcement as implemented in the finite element model is also shown.)
Interface between Concrete and Steel Reinforcement
Figs. 4.16 and 4.17 show the variation of bond stress and bond slip ,respectively, for interface
elements at location I1 (see Fig. 4.2). At the face of the center column a change in sign of bond
stress and bond slip values, compared to the remaining interface elements along the beam, is
observed. Steel reinforcement in compression is pushed into the center block. In order to sat-
isfy the compatibility, the first element of the beam steel reinforcement has to move into the
center block yielding the same sign of bond stress and bond slip values as for interface elements
within the center column. After a transition, bond stress and bond slip variation is consistent
with the change in bending moment along the beam. The same pattern of bond stress and bond
slip variation can be observed for interface elements around steel reinforcement in tension in
the vicinity of the side support. As opposed to steel reinforcement in compression, the an-
chored steel reinforcement in tension is pulled out of the side support. Again, for compatibility
reasons, bond stress and bond slip values of interface elements around the first steel reinforce-
ment element of the beam have the same sign as bond stress and bond slip values of interface
elements within the side support. The maximum bond stress of τmax = 1000 psi is calculated
at a vertical displacement of about 1.68 in at the vicinity of the center column.
ANALYTICAL RESULTS AT MATERIAL LEVEL 45
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−8.5
−7.5
−6.5
−5.5
−4.5
−3.5
−2.5
−1.5
−0.5
0.5
1.5
2.5
3.5
4.55
x 10−3
DISTANCE ALONG BEAM [in]
AV
ER
. ELE
M. R
EL.
DIS
PLA
CE
ME
NT
[in]
I1 Slip Distribution
LS7LS8LS9
Faceof
CenterColumn
Faceof
SideSupport
Top BarCut−OffLocation
Top BarCut−OffLocation
Figure 4.16: Variation of bond slip at cross sectional location I1 at 1.45, 1.68 and 1.9 in of
vertical displacement
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−1500
−1000
−500
0
500
1000
1300
DISTANCE ALONG BEAM [in]
AV
ER
AG
E E
LEM
EN
T T
RA
CT
ION
[psi
]
I1 Bond Stress Distribution
LS7LS8LS9
Faceof
CenterColumn
Top BarCut−OffLocation
Faceof
SideSupport
Top BarCut−OffLocation
Figure 4.17: Variation of bond stress at cross sectional location I1 at 1.45, 1.68 and 1.9 in of
vertical displacement
ANALYTICAL RESULTS 46
The corresponding bond slip is larger than the yield slip of s1= 0.004 which is consistent with
the implemented bond stress-slip relationship. Bond stress and bond slip values for interface
elements further from the face of the center column remain below maximum bond stress τmax
and yield slip s1 throughout the analysis.
Similar results are obtained for the variation of bond stress and bond slip (see Figs. 4.18 and
4.19) for interface elements at cross sectional location I2 (see Fig. 4.2). As for interface
elements at cross sectional location I1 (see Fig. 4.2), the maximum bond stress of τmax =
1000 psi is calculated only for interface elements around compressive reinforcement at the
vicinity of the face of the side support (Fig. 4.19). Also, the corresponding bond slip is larger
than 0.004 which is again consistent with the implemented bond stress-slip relationship. Bond
stress and bond slip values for interface elements more far from the side support remain below
maximum bond stress τmax and yield slip s1, respectively, throughout the analysis. However,
as opposed to interface elements at cross sectional location I1, the maximum bond stress τmax
is calculated to occur at a vertical displacement of about 2.13 in.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.025
−0.02
−0.015
−0.01
−0.005
0
0.005
0.01
0.015
DISTANCE ALONG BEAM [in]
AV
ER
. ELE
M. R
EL.
DIS
PLA
CE
ME
NT
[in]
I2 Slip Distribution
LS10LS11LS12
Faceof
SideSupport
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
CenterColumn
Figure 4.18: Variation of bond slip at cross sectional location I2 at 2.1, 2.35 and 2.58 in of
vertical displacement
BOND SLIP AND SECTION DEFORMATION 47
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−1750
−1500
−1000
−500
0
500
1000
1500
DISTANCE ALONG BEAM [in]
AV
ER
AG
E E
LEM
EN
T T
RA
CT
ION
[psi
]
I2 Bond Stress Distribution
LS10LS11LS12
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
CenterColumn
Faceof
SideSupport
Figure 4.19: Variation of bond stress at cross sectional location I2 at 2.13, 2.35 and 2.58 in of
vertical displacement
4.2 Bond Slip and Section Deformation
In this section, computed nodal displacements at selected locations will be evaluated with re-
gard to sectional deformation as well as steel reinforcement pull-out and push-in. Fig. 4.20
shows the cross sections along the beam for which vertical lines, V2 and V3, are defined.
Along these lines nodal displacements were obtained from the program. The designations of
sections, for instance SEC 1 V2, are subsequently used in this section.
Faceof
CenterColumn
Faceof
SideSupport
8.0in 8.0in
SEC 4 SEC 3 SEC 2 SEC 1 V2V3
Figure 4.20: Designation of beam sections
Interface elements were implemented in the finite element model. As a result, the anchored
steel reinforcement is either pulled out (tension) or pushed into (compression) the side support
block or the center column block, respectively. In table 4.2, the calculated maximum values of
steel reinforcement pull-out and push-in at SEC 1 and SEC 4 are listed.
ANALYTICAL RESULTS 48
Table 4.2: Values of steel reinforcement pull-out and push-in at about 5.26 in of vertical dis-
placement
SEC1 SEC4
V2 V3 V2 V3
Pull Out 0.027 in 0.0136 in 0.04 in -
Push In 0.0167 in - 0.0169 in 0.013 in
The maximum value of bar pull-out is calculated for steel reinforcement in tension anchored in
the center column block (SEC 4 V2). The calculated maximum value of bar pull-out at SEC
1 V2 is 32% less. A comparison of pull-out values calculated for SEC 1 V2 and SEC 1 V3
shows a difference of about 50%. The calculated bar push-in values are within similar range at
both beam end sections.
Figs. 4.21 and 4.22 show nodal displacements at SEC 2 V2 and SEC 3 V2 at a vertical dis-
placement of 1.0 in and 1.9 in. Note that the horizontal axis represents the nodal displacements
in longitudinal direction (z-direction) and the vertical axis represents the section height. As can
be seen, calculated nodal displacements indicate that during the deformation, cross sections re-
main plain if only the nodal displacements of concrete elements are considered. However, as
a result of the bond slip, the cross sections do not remain plain if the nodal displacements of
steel reinforcement elements are also considered.
Figs. 4.23 and 4.24 show nodal displacements at SEC 1 V2 and SEC 4 V2 at a vertical dis-
placement of 1.0 in and Figs. 4.25 and 4.26 show nodal displacements at SEC 1 V2 and SEC
4 V2 at a vertical displacement of 1.9 in. Similar to SEC 2 V2 and SEC 3 V2, the calculated
nodal displacements of the concrete elements of SEC 4 V2 indicate that the section remains
plain throughout the analysis. Yet, as for SEC 2 V2 and SEC 3 V2, due to the bond slip cross
section SEC 4 does not remain plain. The concrete part of SEC 4 shows a non-plain deforma-
tion due to the highly localized pressure at the face of the side support induced by the single
compressive rebar.
BOND SLIP AND SECTION DEFORMATION 49
1
2
3
4
5
6
7
8
-0.01 0 0.01
1
2
3
4
5
6
7
8
-0.01 0 0.01
1
2
3
4
5
6
7
8
-0.01 0 0.01
1
2
3
4
5
6
7
8
-0.01 0 0.01
Figure 4.21: Deformation [in] of SEC
2 at vertical displacements of 1.0 in
(top) and 1.9 in (bottom)
Figure 4.22: Deformation [in] of SEC
3 at vertical displacements of 1.0 in
(top) and 1.9 in (bottom)
ANALYTICAL RESULTS 50
1
2
3
4
5
6
7
8
-0.01 0 0.01
Model: 05_LC21: LoadStep: 5 LNodal TDTXMax/Min onYmax = 7.5Xmax = .84VariationS1_V2
Figure 4.23: Deformation [in] of SEC
1 at vertical displacement of 1.0 in
1
2
3
4
5
6
7
8
-0.01 0 0.01
Figure 4.24: Deformation [in] of SEC
4 at vertical displacement of 1.0 in
1
2
3
4
5
6
7
8
-0.02 -0.01 0 0.01 0.02
Model: 05_23_08_3LC3: Load case 3Step: 9 LOAD: 1Nodal TDTX...G TDTZMax/Min on whole graph:Ymax = 7.5 Ymin = 0Xmax = .139E-1 Xmin = -.144E-1Variation along a lineS1_V2
Figure 4.25: Deformation [in] of SEC 1 at vertical displacement of 1.9 in
ELEMENT END FORCES 51
1
2
3
4
5
6
7
8
-0.01 0 0.01 0.02
Figure 4.26: Deformation [in] of SEC 4 at vertical displacement of 1.9 in
4.3 Element End Forces
In this section, element end forces are presented. In order to calculate the beam internal forces,
nodal reaction forces were obtained at the boundaries of the finite element model (see Fig.
2.5). Figs. 4.27 and 4.28 show beam bending moment response at the boundary of the center
column block and at the boundary of side support block, respectively, versus the imposed
vertical displacement. The horizontal axis represents the vertical displacement imposed at the
boundary of the center column block. The vertical axis represents the moment calculated at the
boundary of the center column and at the boundary of the side support, respectively. The peak
positive bending moment at the boundary of the center column is 175.1 kip-in at a vertical
displacement of 1.9 in (Fig.4.27). The calculated response shows a constantly diminishing
slope up to 1.0 in of vertical displacement. Beyond this level of deformation, the slope of
the bending moment increases. At 1.9 in of vertical displacement the bending moment shows
a drop. Further increased boundary deformations result in decreasing values of the bending
moment. At a vertical displacement of 5.26 in a bending moment of 85.3 kip-in is computed.
ANALYTICAL RESULTS 52
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60
2
4
6
8
10
12
14
16
18x 10
4
Imposed Vertical Displacement [in]
Res
istiv
e M
omen
t [lb
in]
Moment at Center Column Boundary vs. Imposed Vertical Displacement
Figure 4.27: Moment at the boundary of the center column vs. imposed vertical displacement
The peak negative bending moment at the boundary of the side support is 249.9 kip-in at a ver-
tical displacement of 1.68 in (Fig. 4.28). As opposed to the bending moment at the boundary
of the center column, no intermediate increase of the slope of the bending moment is observed.
Instead, the slope of the bending moment is constantly diminishing up to the peak at 1.68 in
of vertical displacement. Increasing boundary deformation beyond vertical displacement of
1.68 in results in a negative slope of the bending moment. At a vertical displacement of 1.9
in, a sharp drop of the bending moment is observed. As boundary deformations are further
increased, the bending moment constantly decreases and a bending moment of 108.8 kip-in is
calculated at a vertical displacement of 5.26 in.
Fig. 4.29 shows the calculated shear force response of the beam. The vertical axis represents the
shear force of the beam and the horizontal axis represents the imposed vertical displacement.
The calculated peak shear force is 4.48 kips at a vertical displacement of 1.9 in. Beyond vertical
displacement of 1.9 in, the shear force reduces as imposed boundary deformations are increased
until maximum vertical displacement of 5.26 in is reached. This level of vertical deformation
corresponds to a shear force of 2.05 kips.
ELEMENT END FORCES 53
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6−2.6
−2.4
−2.2
−2
−1.8
−1.6
−1.4
−1.2
−1
−0.8
−0.6
−0.4
−0.2
0x 10
5
Imposed Vertical Displacement [in]
Mom
ent [
lbin
]
Moment at Side Support Boundary vs. Imposed Vertical Displacement
Figure 4.28: Moment at the boundary of the side support vs. imposed vertical displacement
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60
500
1000
1500
2000
2500
3000
3500
4000
4500
Imposed Vertical Displacement [in]
She
ar F
orce
[lb]
Shear Force at SEC 1 vs. Imposed Vertical Displacement
Figure 4.29: Shear force response vs. imposed vertical displacement
ANALYTICAL RESULTS 54
4.4 State of Strain at Beam Ends
In this section, beam response at the section level is presented. Note that only results at SEC 1
(Face of Side Support) and SEC 4 (Face of Center Column) are discussed (see 4.20). Note that
strain values, used to describe the beam response at section level, are average element strains
of the first elements from either the face of the center column or the face of the side support.
Considering an element length of about 1.0 in, results were calculated at about 0.5 in from the
face of the side support or the the face of the center column. In figures presented, this location
is associated with the first marker (either from the face of the center column or from the face of
the side support) of the curves illustrating the variation of strains. Furthermore, cross sectional
locations defined in Fig. 4.2 are used in figure captions.
The bottom bar at SEC 4 (face of center column) shows a strain of about εs = 0.003 at a vertical
displacement of 0.48 in (see Fig. 4.30, LS3), which is larger than the yield strain. At the same
vertical displacement, the concrete at the top of SEC 4 is at a strain of εc = −0.002 (Fig. 4.31,
LS3)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−1
−0.5
0
0.5
1
1.5
2
2.5
3
3.5x 10
−3
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
S2 Steel Longitudinal Strain Variation
LS1LS2LS3
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut−OffLocation
Top BarCut−OffLocation
Figure 4.30: Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 0.15, 0.3
and 0.48 in of vertical displacement (horizontal dashed line represents yield strain)
STATE OF STRAIN AT BEAM ENDS 55
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−4
−3
−2
−1
0
1
2
3
4
5
6x 10
−3
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
C1 Concrete Longitudinal Strain Variation
LS1LS2LS3
Top BarCut−OffLocation
Faceof
SideSupport
Top BarCut−OffLocation
Faceof
CenterColumn
Figure 4.31: Variation of concrete longitudinal strain εc,zz at cross sectional location C1 at 0.15,
0.3 and 0.48 in of vertical displacement (horizontal dashed lines represent peak compressive
strain and crack strain)
At a vertical displacement of 0.76 in, concrete at the bottom of SEC 1 (face of side support)
shows a strain of about εc = −0.0025 (Fig. 4.32, LS4). The top bar at SEC 1 is at a strain of
εs = 0.002 at this level of vertical displacement which indicates that it has not yielded (Fig.
4.33, LS4).
At 1.2 in of vertical displacement, concrete at the top of the core at SEC 4 is at a strain of
εc = −0.0022 (Fig. 4.34, LS6). Note that concrete at the top at SEC 4 is at a strain of
εc = −0.015 (Fig. 4.35, LS6) which indicates a significant variation of longitudinal strain over
the section height. The top bar at SEC 4 shows a strain of εs = −0.00145 (Fig. 4.33, LS6). The
bottom bar at SEC 4 is at a strain of εs = 0.025 (Fig. 4.36, LS6). Furthermore, at this level of
vertical displacement, concrete at the bottom of the core at SEC 1 is at a strain of εc = −0.0025
(Fig.4.37, LS6). Concrete at the bottom at SEC 1 shows a strain of εc = −0.008 (Fig. 4.32,
LS6). Concurrently, the top bar at SEC 1 has yielded with a strain of εs = 0.0043 (Fig. 4.33,
LS6). The bottom bar at SEC 1 is at a strain of εs = −0.0015 and has not yielded (Fig. 4.36,
LS6).
ANALYTICAL RESULTS 56
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.008
−0.005
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
C4 Concrete Longitudinal Strain Variation
LS4LS5LS6
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
SideSupport
Faceof
CenterColumn
Figure 4.32: Variation of concrete longitudinal strain εc,zz at cross sectional location C4 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent peak compressive
strain and crack strain)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.0015
−0.001
0
0.001
0.002
0.003
0.004
0.005
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
S1 Steel Longitudinal Strain Variation
LS4LS5LS6
Faceof
CenterColumn
Top BarCut−OffLocation
Faceof
SideSupport
Top BarCut−OffLocation
Figure 4.33: Variation of steel longitudinal strain εs,zz at cross sectional location S1 at 0.76, 1.0
and 1.2 in of vertical displacement (horizontal dashed line represents yield strain)
STATE OF STRAIN AT BEAM ENDS 57
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.0025−0.002
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.015
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
DIT
UD
INA
L S
TR
AIN
C2 Concrete Longitudinal Strain Variation
LS4LS5LS6
Faceof
SideSupport
Top BarCut−OffLocation
Faceof
CenterColumn
Top BarCut−OffLocation
Figure 4.34: Variation of concrete longitudinal strain εc,zz at cross sectional location C2 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent peak compressive
strain and crack strain)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.016−0.015
−0.01
−0.005
0
0.005
0.01
0.015
0.017
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
C1 Concrete Longitudinal Strain Variation
LS4LS5LS6
Faceof
CenterColumn
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
SideSupport
Figure 4.35: Variation of concrete longitudinal strain εc,zz at cross sectional location C1 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent peak compressive
strain and crack strain)
ANALYTICAL RESULTS 58
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.002
0
0.005
0.01
0.015
0.02
0.025
0.03
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
S2 Steel Longitudinal Strain Distribution
LS4LS5LS6
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut−OffLocation
Top BarCut−OffLocation
Figure 4.36: Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 0.76, 1.0
and 1.2 in of vertical displacement (horizontal dashed line represents yield strain)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.003
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
C3 Concrete Longitudinal Strain Variation
LS4LS5LS6
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut−OffLocation
Top BarCut−OffLocation
Figure 4.37: Variation of concrete longitudinal strain εc,zz at cross sectional location C3 at 0.76,
1.0 and 1.2 in of vertical displacement (horizontal dashed lines represent peak compressive
strain and crack strain)
STATE OF STRAIN AT BEAM ENDS 59
The top bar at SEC 4 is yielded with a strain of εs = 0.0026 at a vertical displacement of 1.45
in (Fig. 4.38, LS7). Concrete at the top of the core at SEC 4 is at a strain of εc = −0.0057 at
this level of vertical displacement. At a vertical displacement of 1.68 in, the bottom bar at SEC
1 is yielded with a strain of εs = −0.0045 (Fig. 4.39, LS8).
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.009
−0.005
0
0.005
0.01
0.015
0.016
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
S1 Steel Longitudinal Strain Variation
LS7LS8LS9
Top BarCut−OffLocation
Top BarCut−OffLocation
Faceof
SideSupport
Faceof
CenterColumn
Figure 4.38: Variation of steel longitudinal strain εs,zz at cross sectional location S1 at about
1.45, 1.68 and 1.9 in of vertical displacement (horizontal dashed lines represent yield strains)
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90−0.01
−0.005
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
DISTANCE ALONG BEAM [in]
AV
ER
AG
E L
ON
GIT
UD
INA
L S
TR
AIN
S2 Steel Longitudinal Strain Variation
LS7LS8LS9
Faceof
SideSupport
Faceof
CenterColumn
Top BarCut−OffLocation
Top BarCut−OffLocation
Figure 4.39: Variation of steel longitudinal strain εs,zz at cross sectional location S2 at 1.45,
1.68 and 1.9 in of vertical displacement (horizontal dashed lines represent yield strains)
Chapter 5 Comparison between Analytical and
Experimental ResultsIn the chapter, analytical and experimental results are compared. The comparison includes
strains of reinforcing bars and the beam end rotations.
5.1 Steel Reinforcement Strain
Fig. 5.1 compares analytical and experimental strains of the bottom rebar at the face of the cen-
ter column versus the imposed vertical displacement. Note that experimental data are available
only up to a strain of about 0.033.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 20
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
IMPOSED VERTICAL DISPLACEMENT [in]
ST
RA
IN
Analytical StrainExperimental Strain
Figure 5.1: Analytical and experimental strain history of bottom rebar at the face of the center
column (SEC 4) vs. imposed vertical displacement
The analytical strain reaches tensile yield strain of εs = 0.0025 at about 0.48 in of vertical
displacement. The experimental strain reaches tensile yield strain at about 0.68 in of vertical
displacement. After yielding, the slope of experimental strain changes drastically compared
STEEL REINFORCEMENT STRAIN 61
to the slope of analytical strain. This behavior is related to the yield plateau exhibited by the
uniaxial stress-strain relationship of the rebars used in the experimental study. In this region
of the uniaxial stress-strain relationship, strain increases with little or no increase in stress.
Therefore, material stiffness drops considerably leading to large strain increments as vertical
displacement is increased. The yield plateau is not modeled by the uniaxial stress-strain re-
lationship implemented in the model (see Fig. 3.1). This causes significantly different slopes
of the experimental and analytical tensile strains above the yield displacement. After the yield
plateau of the uniaxial stress-strain relationship, stress increases again with increasing strain
yielding enhanced material stiffness. This leads to decreasing tensile strain increments as the
vertical displacement is further increased. Beyond 0.78 in of vertical displacement the slope
of experimental tensile strain approaches the slope of analytical tensile strain. Beyond a ver-
tical displacement of about 1.05 in, the analytical tensile strain is in good agreement with the
experimentally measured tensile strain until a vertical displacement of about 1.6 in is reached.
Fig. 5.2 shows analytical and experimental tensile strains of the top rebar at the face of the
side support vs. the imposed vertical displacement. The analytical tensile strain reaches yield
strain of εs = 0.0025 at about 1.0 in of vertical displacement. The experimental tensile strain
reaches yield strain 1.5 in of vertical displacement. As for the tensile strain history at the face
of the center column, a diminished material stiffness (yield plateau of uniaxial stress-strain
relationship is reached) causes large tensile strain increments in the post yield range of the
experimental strain. After the yield plateau of uniaxial stress-strain relationship, the slope of
experimental tensile strain decreases and the strain increases linearly as vertical displacement is
increased and the analytical tensile strains approach experimental results as the imposed verti-
cal displacement increases. The analytical tensile strains at the face of the center column and at
the face of the side support reach the yield strain at a lower level of vertical displacement com-
pared to the vertical displacement at which experimental tensile strains reach yield strain. The
yield plateau of experimental uniaxial stress-strain relationship is not modeled by the stress-
strain relationship implemented in the finite element model. This leads to a difference between
the slope of the analytical and experimental strains at the vertical displacement level associated
with the yielding strain. However, both analytical tensile strain histories are in good agreement
with experimentally measured tensile strains.
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60
0.005
0.01
0.015
0.02
0.025
0.03
IMPOSED VERTICAL DISPLACEMENT [in]
ST
RA
IN
Analytical StrainExperimental Strain
Bar Fracture(Analytical)
Bar Fracture(Experimental)
Figure 5.2: Analytical and experimental strain history of top rebar at the face of the side support
(SEC 1) vs. imposed vertical displacement
5.2 Beam End Rotations Fig. 5.3 the shows analytical and experimental beam end rotations measured from the face of
the center column over a length of 8 in. Except for the vertical displacements less than about
0.3 in, both experimental and analytical beam end rotations increase almost linearly. Analytical
beam end rotation is in good agreement with measured beam end rotation considering differ-
ence of only about 13% at about 5.26 in of vertical displacement.
Fig. 5.4 compares the analytical and experimental beam end rotations measured from the face
of the side support over a length of 8 in. Beyond a vertical displacement of about 1.0 in a
change in the slope of analytical beam end rotation can be observed. The change of slope in
the experimental beam end rotations starts later at about 1.6 in of vertical displacement. At
about 5.26 in of vertical displacement analytical beam end rotation exceeds experimental beam
end rotation by approximately 20%. As for beam end rotations at the face of center column,
the analytical beam end rotation at the face of the side support is in good agreement with
experimental beam end rotation.
63
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
IMPOSED VERTICAL DISPLACEMENT [in]
RO
TA
TIO
N [r
ad]
Analytical Rotation Face of Center ColumnExperimental Rotation Face of Center Column Bar Fracture
(Analytical)
Bar Fracture(Experimental)
Figure 5.3: Analytical and experimental beam rotations over 8 in from the face of center column
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 60
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
IMPOSED VERTICAL DISPLACEMENT [in]
RO
TA
TIO
N [r
ad]
Analytical Rotation Face of SupportExperimental Rotation Face of Support
Bar Fracture(Analytical)
Bar Fracture(Experimental)
Figure 5.4: Analytical and experimental beam rotations over 8 in from the face of the side
support over a length of 8 in
CONCLUDING REMARKS 64
Chapter 6 Concluding Remarks
The detailed finite element model of the continuous beam developed in this study allowed es-
timating analytical results that in general were in good agreement with the previously obtained
experimental responses of the beam subjected to large deformations. One of the important
response measures in large deformation response of RC beams is the deformation capacity cor-
responding to the first bar fracture. The analytical vertical displacement associated with the
first bar fracture underestimated the corresponding experiential value by about 12%.
The current version of the Diana program does not account for the interaction between the
concrete damage (crushing and cracking) and the bond deterioration. The inclusion of such
an interaction would extend the plastic zone, leading to a more ductile response, which in turn
would reduce the difference between the analytical and experimental vertical displacements
associated with the first bar fracture.
The element rotation is another important and practical measure of deformation response. The
analytical beam rotations over eight inches at the beam ends were in good agreement with the
experimental results. The finite element modeling described in this study can be used in esti-
mating rotation capacities of RC elements.
Bibliography[1] Sasani, M., and Kropelnicki, J. (2007)., ”Progressive Collapse Analysis of an RC Struc-
ture.”, The Structural Design of Tall and Special Buildings, DOI: 10.1002/tal.375.
[2] Popov, E., P. (1991)., ”Engineering Mechanics of Solids”,Prentice-Hall Inc.
[3] Chen, W., F., and Han, D., J. (1988)., ”Plasticity for Structural Engineers”, Springer-Verlag
New York Inc.
[4] DIANA Finite Element Analysis (2007)., Release 9.2, Users’s Manual, Material Library.,
DIANA TNO, BV.,
[5] Park, R. (1975)., ”Reinforced Concrete Structures”, John Wiley & Sons Inc.
[6] Rots, J.G., and Blaauwendraad, J. (1989)., ”Crack Models for Concrete: Discrete or
Smeared? Fixed Multi-Directional or Rotating?”, Heron, Vol. 34, p. 3-59
[7] Vecchio, F. J., and Collins, M. P. (1986)., ”The Modified Compression Field Theory for
Reinforced Concrete Elements Subjected to Shear.”, ACI Journal 83(22), p. 219-231.
[8] R. G. Selby, and F. J. Vecchio (1993)., ”Three-dimensional Constitutive Relations for
Reinforced Concrete Reinforced Concrete.”, Tech. Rep. 93-102, Univ. Toronto, Dept. Civil
Eng., Toronto, Canada
[9] Collins, M., and Porasz, A. (1989)., ”Shear Design for High Strength Concrete”,
ComiteEuro-International du Beton, Bulletin d’ Information No. 193, p. 77-83
[10] Gilbert, R., I. and Warner, R., F. (1978). ”Tension Stiffening in Reinforced Concrete
Slabs.” Journal of Structural Engineering, Vol. 104(12) p. 1885-1900
[11] Mander, J., B., Priestley, M., J., M. and Park, R. (1988)., ”Theoretical Stress-Strain Model
for Confined Concrete.”, Journal of Structural Engineering, Vol 114(8), p. 1804-1825
[12] Eligehausen, R., Popov, E., P. and Bertero, V., V. (1983)., ”Local Bond Stress-Slip Re-
lationships of Deformed Bars under Generalized Excitations.”, Report No. UCB/EERC-
83/23, University of Berkeley
BIBLIOGRAPHY 66
[13] FIB- Federation Internationale du Beton (2000)., ”Bond of Reinforcement in Concrete.”,
Chapter 1: ”Bond Mechanics Including Pull-out and Splitting Failure.”, (principal authors
Gambarova, P., G., Plizzari, G., A., Rosati, G., P. and Russo G.)
[14] Tepfers, R. (1979)., ”Cracking of concrete cover along anchored deformed reinforcing
bars.”, Magazine of Concrete Research, Vol.32(106), p. 3-12
[15] Nilson, A. H. (1971)., ”Bond Stress-Slip Relations in Reinforced Concrete.”, Report 345,
School of Civil and Environmental Engineering, Cornell University, Ithaca
[16] Fardis, M. W and Buyukozturk, O. (1980)., ”Shear Stiffness of Concrete by Finite Ele-
ments.”, ASCE Journal of the Structural Division, Vol. 101(ST6), p. 1311-1327
[17] Mirza, S., M. and Houde, J. (1978)., ”Study of Bond Stress-Slip Relationships in Rein-
forced Concrete.”, Journal of The American Concrete Institute, Vol 76(1), 978pp